Regulation of EMT by TGFβ in cancer

FEBS Letters 586 (2012) 1959–1970

journal homepage: www.FEBSLetters.org

Review

Regulation of EMT by TGFb in cancer Carl-Henrik Heldin a,⇑, Michael Vanlandewijck a, Aristidis Moustakas a,b a b

Ludwig Institute for Cancer Research, Uppsala University, Box 595, SE-751 24 Uppsala, Sweden Department of Medical Biochemistry and Microbiology, Science for Life Laboratory, Uppsala University, Box 582, SE-751 23 Uppsala, Sweden

a r t i c l e

i n f o

Article history: Received 13 January 2012 Revised 21 February 2012 Accepted 21 February 2012 Available online 28 February 2012 Edited by Joan Massagué and Wilhelm Just Keywords: Cancer stem cell Epithelial-mesenchymal transition MAPK Metastasis Smad TGFb

a b s t r a c t Transforming growth factor-b (TGFb) suppresses tumor formation since it inhibits cell growth and promotes apoptosis. However, in advanced cancers TGFb elicits tumor promoting effects through its ability to induce epithelial-mesenchymal transition (EMT) which enhances invasiveness and metastasis; in addition, TGFb exerts tumor promoting effects on non-malignant cells of the tumor, including suppression of immune surveillance and stimulation of angiogenesis. TGFb promotes EMT by transcriptional and posttranscriptional regulation of a group of transcription factors that suppresses epithelial features, such as expression of components of cell junctions and polarity complexes, and enhances mesenchymal features, such as production of matrix molecules and several cytokines and growth factors that stimulate cell migration. The EMT program has certain similarities with the stem cell program. Inducers and effectors of EMT are interesting targets for the development of improved diagnosis, prognosis and therapy of cancer. Ó 2012 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved.

1. Introduction The transforming growth factor-b (TGFb) family of cytokines has 33 members in humans, including TGFb isoforms, activins, bone morphogenetic proteins (BMPs), and growth and differentiation factors (GDFs). These factors regulate growth, survival, differentiation and migration of cells, and have important roles during embryonal development and in the control of adult tissue homeostasis [1,2]. TGFb family members are implicated in several diseases, including fibrotic conditions, autoimmune diseases and cancer [3]. During carcinogenesis, TGFb has a dual role; initially it suppresses tumorigenesis by inducing growth arrest and promoting apoptosis, however, in advanced cancers, where TGFb often is overexpressed [4], it promotes tumorigenesis by induction of epithelial-mesenchymal transition (EMT), whereby tumor cells become more invasive and prone to form metastases [5]. The tumor promoting effects of TGFb also include effects on non-malignant cells of the tumor; thus, TGFb suppresses immune surveillance, and promotes angiogenesis and the recruitment of inflammatory cells which secrete several cytokines that act on the tumor cells (Fig. 1). TGFb signals via formation of a heterotetrameric complex of type I and type II serine/threonine kinase receptors, in which the constitutively active type II receptor phosphorylates and activates the type I receptor [1,2]. A crucial signaling pathway downstream of ⇑ Corresponding author. E-mail address: [email protected] (C.-H. Heldin).

TGFb receptors is composed of transcription factors of the Smad family (Fig. 2). Receptor-activated Smads (Smad2 and 3 for TGFb) are phosphorylated in SXS-motives in their C-terminals, whereafter they form complexes with the common-mediator Smad4. Smad complexes accumulate in the nucleus where they, in cooperation with other transcription factors, co-activators and co-repressors, regulate the transcription of specific genes (Fig. 2). Smad signaling is carefully regulated by posttranslational modifications of Smads involving activating and inhibitory phosphorylations by different kinases, and ubiquitination by different ubiquitin ligases that control stability and activity of Smads. One of the target genes of TGFb is the Smad7 gene; induction of the inhibitory Smad7 creates a negative feedback loop which controls TGFb signaling. In addition to Smad signaling, TGFb induces several other signaling pathways, e.g. the Erk MAP kinase pathway which is activated via tyrosine phosphorylation of the adaptor Shc, allowing docking of the Grb2Sos1 complex which leads to activation of Ras and the downstream Erk MAP kinase pathway [6]. In addition, the ubiquitin ligase TRAF6 binds to the TGFb type I receptor and mediates TGFb activation of the MAP kinase kinase kinase TAK1, the MAP kinase kinase 3/6, and the Jun/p38 MAP kinases (Fig. 2) [7,8]. TRAF6 also mediates cleavage of the type I TGFb receptor in a manner which is dependent on the metalloproteinase TACE and protein kinase C (PKC), whereafter the intracellular part of the receptor is translocated into the nucleus where it induces a transcriptional program promoting cell invasiveness (Fig. 2) [9]. Additional non-Smad pathways induced by TGFb include phosphorylation of PAR6 by the TGFb type II

0014-5793/$36.00 Ó 2012 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.febslet.2012.02.037

1960

C.-H. Heldin et al. / FEBS Letters 586 (2012) 1959–1970

Fig. 1. Actions of TGFb during cancer progression. A growing primary carcinoma along the epithelial layer of an organ is shown on top with the lumen of the organ and the underlying basement membrane. Round pink cells represent hyperproliferating benign cells. Round red cells represent more advanced carcinoma cells that then undergo EMT (triangular red cells) and move towards the blood vessel that is surrounded by yellow endothelial cells. Various stromal cell components are represented by elongated beige (e.g. cancer associated fibroblasts) and polygonal green cells (e.g. macrophages). The bottom part captures an aspect of extravasation and metastatic colonization, with most cells shown as round red cells, emphasizing MET at the metastatic site. Word captions summarize key events during the progression.

receptor which inactivates the epithelial polarity complex (Fig. 2), as well as activation of the tyrosine kinase Src, the small GTPase Rho, and the phosphatidylinositol 30 -kinase (PI3-kinase) [1]. TGFb was first shown to induce EMT almost 20 years ago [10]. The aim of the present review is to discuss the mechanism whereby TGFb induces EMT and its role in TGFb-induced invasiveness and metastasis of tumor cells. 2. Characteristics of EMT EMT is typically characterized by the loss of cell–cell adhesions and apical-basal polarity, and the development of a fibroblastoid motile phenotype. Together with its reversal, mesenchymal-to-epithelial transition (MET), EMT occurs normally during embryonal development, e.g. in the formation of mesenchyme during gastrulation, in neural crest delamination, in the development of placenta, somites, heart valves, urogenital tract and secondary palate, as well as during branching morphogenesis of different organs [5]. EMT is characterized by loss of epithelial cell-cell contacts by suppression

of components that build up junctional complexes, such as E-cadherin, ZO-1, CAR, occludin, claudin-1 and claudin-7 [1]. Cells hereby become capable of detaching from each other and move away from the organized epithelial tissue. Cells also lose polarity via loss of components of polarity complexes, e.g. CRB3 and LGL2. Changes in microfilaments, microtubules and intermediate filaments further promote the architectural reorganization and cell motility. In parallel, cells acquire mesenchymal features, such as production of matrix proteins, e.g. fibronectin and collagen, and production of metalloproteases, which also facilitates cell migration. Finally, a number of cytokines and growth factors are induced during EMT, which amplifies the EMT program and promotes cell migration, including the cytokines TGFb, Wnt, Notch ligands, ILEI (Interleukin-like EMT-inducer), HGF (hepatocyte growth factor), EGF (epidermal growth factor) and PDGF (platelet-derived growth factor). In addition to EMT, there are also examples of other reprogramming processes, e.g. endothelial-mesenchymal transition (endoMT) and the differentiation of epithelial-derived mesenchymal cells to myofibroblasts (EMyoT).

C.-H. Heldin et al. / FEBS Letters 586 (2012) 1959–1970

1961

Fig. 2. TGFb receptor signaling and the EMT program. Two adjacent epithelial cells are shown tethered via: (a) tight junctions made of occludin (occ), claudins (clau) and JAM proteins that are bound to ZO family proteins that crosslink to the actin fibers (red crossed lines); (b) adherence junctions made of E-cadherin (E-cad) to which is bound bcatenin (b-cat) and a-catenin (a-cat). Four different scenarios of TGFb receptor tetrameric complexes are shown providing different signaling outputs (from left to right): (i) TGFb receptor signaling to Smads, with R-Smad phosphorylation followed by complex formation with the Co-Smad, nuclear accumulation of the complex and binding to DNA (double helix) together with transcription factors (TF). (ii) TGFb receptor signaling to p38 MAPK via poly-ubiquitination (blue circles) of TRAF6 and TAK1, and phosphorylation of MKK3/6 and p38. Smad7 serves as a platform for the kinases. MAPK signaling feeds to the transcriptional complex made by Smads and TFs. (iii) The TGFb receptor complex with TRAF6 activates the protease TACE that cleaves the type I receptor and releases its intracellular domain (ICD) that moves to the nucleus and binds to chromatin. These three signaling pathways regulate transcription of genes leading to downregulation (red small circle) of the epithelial program and upregulation (green small circle) of the mesenchymal program. (iv) TGFb receptor signaling to the polarity complex via receptors localized at tight junctions, with the type II receptor phosphorylating Par6, leading to recruitment of the ubiquitin ligase Smurf1 that degrades RhoA. The signaling and junctional complexes are shown in different cells for clarity, but are of course present in all epithelial cells.

3. EMT in disease EMT is promoted by chronic inflammation and occurs in conjunction with tissue fibrosis [11], during which epithelial cells are converted to myofibroblasts, the latter secreting large amounts of collagen and other matrix molecules, as revealed by cell tracking studies [12]. Evidence has been obtained that EMT is important during CCL4-induced liver fibrosis [13] and TGFb-induced lung fibrosis [14]. During cardiac [15] and kidney [16] fibrosis, endothelial cells are converted to mesenchymal cells by Endo-MT. The profibrotic effect of TGFb, e.g. its well-established ability to promote accumulation of matrix molecules by inducing their synthesis and inhibiting their degradation, reflects its EMT promoting activity. In several tumor models in animals, it has been demonstrated that EMT occurs and promotes invasiveness and metastasis. Direct evidence that EMT is of relevance in human cancers has been more difficult to obtain, possibly due to the fact that EMT is transient and reliable markers have been lacking [5]. Evidence supports the notion that tumor invasion may be aided by EMT at the invasive front of tumors [17], but tumors can also invade their environment on an EMT-independent manner [18]. Pathological EMT is similar to physiological EMT and involves similar signaling pathways (see further below).

4. Mechanism of TGFb induction of EMT At the heart of TGFb regulation of EMT is a nuclear reprogramming involving a set of transcription factors, i.e. the basic helix loop helix proteins Twist and E47, the zinc finger proteins Snail and Slug

(also called Snail2), the zinc finger and homeodomain proteins ZEB1 (also called dEF1) and ZEB2 (also called SIP1) [19], and FOXC2 [20]. These factors regulate each other in an elaborate manner (Fig. 3). Thus, Snail upregulates Slug [21,22] and Twist [23], Snail and Twist induce ZEB1 [24,25] and Slug [26], and Snail induces ZEB2 [22]. Together, the factors repress the expression of E-cadherin, which is a crucial step of EMT, as well as other epithelial markers, but also enhance the expression of mesenchymal genes. As will be discussed below, several pathways induced by TGFb contribute to the regulation of these transcription factors. Moreover, TGFb cooperates with several other factors such as Notch, Wnt, integrins and certain tyrosine kinase receptor ligands, in their regulation. 4.1. Smad signaling is important for TGFb-induced EMT TGFb-induced activation of Smad complexes has crucial roles during induction of EMT [27–30]. However, whereas Smad4 and Smad3 promote EMT, Smad2 inhibits EMT of keratinocytes [31]. Given the important roles of Smad molecules in TGFb-induced EMT, it is not surprising that mechanisms that affect Smad levels or activities affect EMT. Thus, ablation of Smad2 in keratinocytes enhances EMT of keratinocytes [31] and of human renal epithelial cells [32]. In addition, ubiquitin ligases that promote poly-ubiquitination and proteasomal degradation of Smads, such as WWP2, have been shown to affect EMT [33]. Moreover, the transcription intermediary factor 1c (TIF1c), which can ubiquitinate Smad4 [34] or replace Smad4 in a Smad2/3 complex [35], was shown to suppress TGFb-induced EMT [36]. Smad signaling induces the expression of high mobility group A2 (HMGA2), a nuclear protein which contains three AT-hooks

1962

C.-H. Heldin et al. / FEBS Letters 586 (2012) 1959–1970

of the receptor binds cooperatively with Smad complexes to the promoters of genes involved in the invasiveness program. 4.4. Epigenetic mechanisms control EMT

Fig. 3. The nuclear program of EMT induced by TGFb stimulation. TGFb signaling is represented by a box that summarizes the three pathways (i–iii) presented in Fig. 2. The nuclear program is divided in two parts: (a) Smad, MAPK and TGFb type I receptor ICD signaling leads to induction of the key transcription factors HMGA2, Snail, Slug, Twist and ZEB. HMGA2 induces expression of Snail and Twist. Snail induces expression of Slug, Twist and ZEB. Twist induces expression of Slug and ZEB. Snail, Slug, Twist and ZEB regulate the expression of genes of the epithelial and mesenchymal program. Red small circles show negative regulatory connections, while green small circles show positive regulatory connections. (b) The transcriptional module also induces expression of miR-155 that negatively regulates RhoA; of miR-491-5 that negatively regulates Par3; of Twist that induces miR-10b, which negatively regulates the homeobox factor HOXD10; and of ZEB, which transcriptionally represses miR-200 family genes that negatively regulate SIRT1 and ZEB. Finally, TGFb signaling (extended arrow representing non-Smad signaling pathway) activates Akt2, which leads to transcriptional repression of the miR-200 family. In addition, hypoxia is shown to induce HIF-1a, which binds to HDAC3 and leads to activation of Snail and Twist, in a cooperative manner with TGFb signaling.

which binds to the minor grove of DNA and interacts with several other transcription factors, including Smads. HMGA2 upregulates Snail [21,22] and Twist [37]. Interestingly, HMGA2 cooperates with Smads in the induction of Snail in a feed-forward mechanism (Fig. 3). 4.2. Smads cooperate with other pathways in TGFb-induced EMT TGFb has been shown to cooperate with oncogenic Ras in the induction of Snail and promotion of EMT [38]. In Ras-transformed breast cancer cells, TGFb upregulates the transcription factor ATF3, which in turn induces Snail, Slug and Twist [39]. Interestingly, the cooperation between TGFb and Ras is further enhanced by activation of the transcription factor NF-jB pathway which enhances Snail expression [40]. The important topic of signaling crosstalk is revisited in further detail below. 4.3. Non-Smad pathways contribute to TGFb-induced EMT A non-Smad signaling pathway which may be of importance in the promotion of EMT is the TRAF6-mediated cleavage of the type I receptor by the metalloproteinase (MMP) TACE [9]. The intracellular part of the receptor is translocated into the nucleus where it drives the expression of genes involved in tumor cell invasiveness, e.g. the Snail and MMP2 genes (Fig. 2). An interesting possibility, which remains to be elucidated, is that the intracellular domain

TGFb has been shown to induce the DNA methyl-transferase DNMT1 in a Smad-dependent manner [41], which contributes to an epigenetic control of EMT. Moreover, posttranslational modifications of histone 3 regulate conformation of chromatin at specific genomic regions and thereby promote EMT; thus, TGFb stimulation causes a reduction in the heterochromatin mark di-methylation of Lys9, an increase in the euchromatin mark tri-methylation of Lys4, and an increase of the transcriptional mark tri-methylation of Lys36 in histone 3 [42]. These changes depend mainly on lysinespecific demethylase-1 (Lsd1), and knock-down of Lsd1 by siRNA inhibits EMT [42]. A recent interesting study showed how hypoxia can enhance EMT via an epigenetic mechanism, possibly in cooperation with TGFb stimulation. Hypoxia causes upregulation of Snail and Twist, as well as of HIF-1a (Fig. 3). The latter upregulates the histone deacetylase HDAC3, which deacetylates Lys4 of histone 3, leading to a repression of epithelial-associated gene chromatin structure, which, in conjunction with Snail and Twist, lower the expression of epithelial genes [43]. In parallel, in mesenchymal cells, HDAC3 collaborates with the histone methyltransferase WDR5 which methylates Lys4 of histone 3, causing enhanced expression of mesenchymal genes [43]. Through this dual epigenetic mechanism, HDAC3 efficiently promotes EMT. Interestingly, Smad3 has been shown to collaborate with another histone deacetylase, i.e. HDAC6, in the induction of EMT [44]. HDAC6 deacetylates a-tubulin in microtubules and thereby affects cell motility. 4.5. Loss of polarity is an important part of EMT TGFb causes loss of polarity of epithelial cells by lowering the amounts of the Par3 protein, which results in a redistribution of the Par6-aPKC complex from the plasma membrane to the cytoplasm and its subsequent disorganization [45]. In addition, Par6 interacts with the type I TGFb receptor; upon TGFb stimulation Par6 is phosphorylated by the type II receptor, leading to recruitment of the ubiquitin ligase Smurf1 to the junctional complex (Fig. 2). Hereby, the GTPase RhoA is targeted for ubiquitination and proteasomal degradation, which promotes EMT [46]. Another mechanism whereby TGFb induces loss of polarity is via Twistmediated induction of ZEB1 [37], which represses the polarity protein LGL2 [47]. 4.6. miRNAs are involved in TGFb-induced EMT The levels of transcription factors driving EMT are controlled by miRNAs (Fig. 3). The miR-200 family (miR-200a, miR-200b, miR200c, miR-141 and miR-429) and miR-205 have particularly important roles since they target ZEB1 and ZEB2 mRNAs [48–51]. TGFb downregulates all five members of the miR-200 family via activation of the Akt2 serine/threonine kinase [52], thereby allowing accumulation of ZEB1 and ZEB2. The expression of ZEB2 is further regulated by a naturally occurring antisense transcript [53]. Interestingly, ZEB2 binds to the promoters of the miR-200 members and represses their expression, thus constituting a doublenegative regulatory loop [48,54]. The miR-200 family also targets the histone deacetylase SIRT1 which further contributes to EMT by affecting the epigenetic regulation of EMT genes [55]. TGFb also upregulates certain miRNAs, including miR-155, which targets RhoA [56]. Together with the Par6 mechanism described above, this causes a decrease in RhoA level, which pro-

C.-H. Heldin et al. / FEBS Letters 586 (2012) 1959–1970

motes dissolution of tight junctions and loss of polarity. This effect is enhanced by another miRNA which also is induced by TGFb, i.e. miR-491-5, which targets the polarity protein Par3 [57]. TGFb also upregulates miR-10b, via induction of Twist; miR-10b targets HOXD10 and enhances invasiveness and metastasis of breast cancer cells [58]. Several other miRNAs have also been found to be up- or downregulated during EMT, suggesting that EMT is controlled by a network of miRNAs; however, their specific roles to promote or control TGFb-induced EMT remain to be elucidated. 4.7. TGFb-induced differential splicing promotes EMT TGFb has been shown to induce alternative splicing in breast epithelial cells, whereby fibroblast growth factor (FGF) receptors 1, 2 and 3 are converted from the IIIb forms to the IIIc forms [59]. This causes a switch in ligand binding specificities from FGF-7 and FGF-10, to FGF-2 and FGF-4; FGF-2 then cooperates with TGFb to induce EMT. TGFb affects splicing by ZEB1- and ZEB2-mediated downregulation of epithelial splicing regulatory proteins (ESRPs) [60]. ESRPs also affect the splicing of many other genes and their downregulation contributes significantly to EMT [61,62]. 4.8. Regulation of mRNA translation contributes to TGFb-induced EMT As described above, TGFb stimulation leads to activation of Akt2, which phosphorylates the ribonucleoprotein E1 (hnRNP1 E1). hnRNP E1 in complex with the elongation factor 1A1 (eEF1A1) binds to a 33 nucleotide sequence in the 30 untranslated regions of the mRNAs for the adaptor protein Dab2 and the cytokine ILEI, and represses their translation; Akt2-mediated phosphorylation of hnRNP E1 dissociates the complex and allows translation of the mRNAs [63,64]. The importance of this mechanism for EMT was further consolidated by the finding that knock-down of hnRNP E1 expression enhances EMT and metastasis. 5. EMT as a product of crosstalk between signaling pathways TGFb is a particularly efficient inducer of EMT, but EMT is also promoted by a number of other factors, including Notch, Wnt, IL6 and several tyrosine kinase ligands, such as FGF, EGF, PDGF and HGF. Interestingly, members of the BMP family instead inhibit EMT and promote MET, however, this happens in a tissue- or context-dependent manner as explained below. Exactly how all these factors contribute to EMT remains to be elucidated, however, tyrosine kinase receptor-induced activation of the Ras-Erk MAP kinase pathways is likely to act synergistically with TGFb and to contribute to multiple aspects of the EMT, including pro-invasive and prometastatic behavior of tumor cells of diverse tissue origins [65]. Moreover, several of these growth factors and cytokines induce each other, thus amplifying and sustaining the EMT process. In the following sections we discuss selected examples of crosstalk between TGFb and other growth factors, focusing on the signaling mechanisms that mediate such crosstalk. We attempt to emphasize the context-dependency meaning that sometimes one and the same growth factor contributes to opposite biological outcomes (e.g. EMT or MET). 5.1. Wnt and TGFb crosstalk The Wnt pathway is implicated in both developmental and tumor-related EMTs. In invasive breast cancer models, canonical Wnt signaling involves the mobilization of b-catenin to the nucleus where it acts as a transcriptional cofactor of TCF (T-cell factor) [66]. The mechanism whereby this pathway elicits EMT involves

1963

the sequestration of the kinase GSK-3b by the adaptor protein Axin2, thus prohibiting the kinase from actively phosphorylating the transcription factor Snail and causing its degradation. This leads to Snail stabilization, allowing cooperation between Snail and TCF in promoting the EMT phenotype [66]. Interestingly, the basic b-catenin/TCF or LEF (lymphoid enhancer factor) signaling module is often usurped by TGFb signaling in several tissues to elicit EMT, presumably without any direct involvement of Wnt ligands. Accordingly, developmental EMT that forms the palate under the signaling activity of TGFb3 involves direct association of the Smad complex with LEF1, which represses E-cadherin gene expression [67]. Thus, TGFb signaling induces expression of Snail and Wnt signaling stabilizes Snail; since Smads can bind to both LEF1 and Snail [68], it is possible that during Wnt and TGFb-induced EMT, a common Smad-LEF-Snail complex forms to promote EMT. During pulmonary fibrosis, lung epithelial cells exhibit a cell surface receptor complex between TGFb receptors, integrin receptor a3b1 and E-cadherin, which allows the formation of signaling complexes between Smads and b-catenin that promote EMT [69]. The Smad/b-catenin complex induced by TGFb in lung epithelial cells, together with the co-activator p300, also induce the mesenchymal a-smooth muscle actin gene [70]. TGFb also induces the expression of the homeobox transcription factor CUTL1, which then induces the expression of the ligand Wnt-5A, activating canonical Wnt signaling and promoting EMT in pancreatic carcinomas [71]. In addition to canonical Wnt signaling, TGFb can induce and cooperate with non-canonical Wnt signaling and together these three signaling pathways form a network that promotes EMT in mammary cells [72]. The regulated secretion of extracellular antagonists of both TGFb/activin and Wnt ligands help maintain the epithelial phenotype by protecting the mammary cell from being exposed to autocrine TGFb or Wnt signals. When the EMT process starts by pathological exposure to high levels of these ligands, the extracellular inhibitors become downregulated, leading to sustained autocrine responses to TGFb and Wnt, promoting the development of transitory mesenchymal cells with tumor-initiating capacities (see also next section). 5.2. Notch and TGFb crosstalk The Notch pathway that specifies cell fate determination during development crosstalks significantly with the TGFb pathway [73]. Notch catalyzes EMT during organogenesis, such as heart valve formation and during oncogenesis [74]. The primordial cells that form the heart valves are produced via an Endo-MT during which Notch signaling induces expression and secretion of TGFb, which then induces Snail expression and downregulation of vascular-endothelial (VE) cadherin. In the same biological system, Notch signaling leads to Slug expression, whereas Snail expression requires the concerted action of Notch and TGFb [75]. Interestingly, Notch and TGFb cooperate with BMP2 in the induction of Endo-MT in the endocardial tissue [76]. BMP2, similar to Wnt, inhibits GSK3b activity, thus leading to Snail stabilization. Thus, EMT driven by Snail requires both transcriptional induction of its gene, e.g. via TGFb signaling, but also additional stabilization of the protein via Wnt or BMP signaling, depending on the cell type that undergoes the transition. In cultured cell models of EMT, such as human HaCaT keratinocytes, human kidney epithelial HK-2 cells and mouse NMuMG mammary epithelial cells, TGFb signaling, via Smads, leads to expression of several Notch receptor ligands, including Jagged1 [77–79]. Jagged1 then induces paracrine signaling of the Notch pathway so that effective epithelial cell cycle arrest and EMT can take place. Interestingly, genome-wide screens for genes regulated by TGFb and Notch, demonstrated that a large cohort of genes are co-regulated by the action of these two pathways. This mechanism

1964

C.-H. Heldin et al. / FEBS Letters 586 (2012) 1959–1970

of crosstalk is also relevant to human cancer as demonstrated in the case of a subset of renal cancer where the TGFb/Notch pathway signature provides a tool useful for the prognosis of disease progression [80]. Moreover, Notch signaling induces secretion of TGFb, as shown in lung epithelial cells undergoing EMT and terminal myofibroblast differentiation [81]. An important marker of myofibroblast differentiation, a-smooth muscle actin, is induced in the lung cells in response to Notch activation in a TGFb- and Smad3-dependent manner, while the Jagged1 ligand is also induced so that a sustained Notch pathway activity can be elicited. On the other hand, in different cell types, such as esophageal epithelial cells that develop squamous carcinomas, Notch signaling by the Notch3 receptor counteracts EMT by affecting the balance of the miR-200/ZEB/TGFb axis described above, and thus promoting epithelial differentiation [82]. 5.3. PDGF and TGFb crosstalk The importance of oncogenic Ras and TGFb signaling for EMT and metastasis has been summarized above. The studies on the Ras-TGFb crosstalk also revealed that an important factor induced after TGFb signaling is PDGF that acts in a sequential auto- or paracrine manner to promote the sustained EMT and cancer cell invasiveness, as first demonstrated in liver cancer cells [83]. In fact, both PDGF-A and the a- and b-receptors are transcriptionally induced by TGFb signaling in Ras-transformed hepatocarcinomas. In this tumor type, activation of PDGF signaling after prolonged TGFb exposure leads to additional intracellular signaling via the PI3K and b-catenin routes that enhance carcinoma survival and invasiveness [84]. In addition, the paracrine action of the TGFb-induced PDGF was shown to target stromal myofibroblasts, which support the invasive phenotype of the hepatocarcinomas [85]. The input of TGFb signaling on PDGF secretion and PDGFR expression may not be direct, as recent analysis of the TGFb–PDGF crosstalk in hepatocarcinomas and in lung carcinomas revealed that the TGFb-inducible secretion of ILEI (see above for mechanism) is an important mediator of PDGF secretion and PDGF receptor expression, which then lead to signaling via b-catenin and Stat3 to establish EMT [86]. The TGFb–PDGF crosstalk mechanism was also established in metastatic breast cancer cells, whereby both PDGF a- and b-receptors were upregulated in response to TGFb, and the resulting PDGF signals contributed to EMT and invasive growth by stimulation of several downstream pathways including the Stat1 pathway [87]. The extensive crosstalk between multiple signaling pathways in the induction of EMT indicate that efficient inhibition of EMT during tumor progression using pharmacological agents, may need to involve the targeting of cooperating signaling pathways simultaneously. 6. Links between EMT and tumor-initiating cells Over the past few years, a new concept of cancer progression has been explored. It emphasizes the importance of a relatively rare cancer cell type with properties similar to stem cells, that contributes to the growth, expansion and dissemination of the tumor and also to its relative resistance to chemotherapy [88]. In reference to this rare cell type, we use the term tumor-initiating cell (TIC) instead of cancer stem cell, since it is unlikely that cancer cells can fully de-differentiate to generate genuine stem cells that can then repopulate and reconstitute a normal tissue, organ or a complete organism. The cellular and molecular features of TICs and cancer cells undergoing EMT show interesting similarities. Thus, transcription factors of the EMT program, such as Snail and Twist, but also cytokines that promote EMT, such as TGFb, promote

the generation and survival of TICs [72,89,90]. A prominent cell type with potential to promote tumorigenesis and assist tumor cells to undergo the transition towards metastasis, are mesenchymal stem cells (MSCs). It is therefore interesting that overexpression of Snail or Twist in breast cancer cells produces cancer cell phenotypes with common molecular and behavioral features with the MSCs [91]. We explained above how Snail and Twist expression is controlled by growth factor pathways, including TGFb. Under in vitro culture conditions of normal fetal liver cells, TGFb signaling leads to Snail induction and EMT that produces mesenchymal cells with progenitor-like phenotypic features [92]. Furthermore, TGFb and Wnt signaling contribute to the emergence and survival of mesenchymal cells with progenitor-like phenotype and enhanced tumorigenic potential from breast cancer cells [72]. Moreover, TGFb and tumor necrosis factor-a (TNFa) induce EMT and generation of TICs of breast cancer cells that exhibit resistance to anticancer drugs [93]. The mesenchymal cells produced via EMT lack their tight junctions, as explained above, and exhibit low levels of expression of proteins that assemble the tight junction, such as the claudins. In particular, claudin 3, 4 and 7 are downregulated in breast cancer EMT and this phenotype is now used for the classification of a subtype of breast cancers as ‘‘claudin-low’’ [93]. In addition to the similarities between EMT and the generation of TICs, there are similarities between MET and the generation of induced pluripotent stem cells (iPSCs) from somatic cells, such as fibroblasts. The molecular process that induces the epithelial phenotype during the generation of iPSCs is controlled by the embryonic stem cell transcription factors Oct4, Sox2, Klf4 and the proto-oncoprotein c-Myc. Oct4 and Sox2 bind to and repress the Snail gene expression, Klf4 induces expression of E-cadherin driving the epithelial differentiation of iPSCs, and the proto-oncogene c-Myc represses the genes encoding TGFb and the type II TGFb receptor [94]. It should be pointed out that all studies so far that link the process of EMT with TICs in human and experimental cancers rely on in vitro culture protocols of tumor cells and their genetic or growth factor-mediated manipulation. The relevance of these findings for the actual tumors in patients requires further research. At the molecular level, we need to understand how pathways like TGFb and transcription factors like Snail change the phenotype of tumor cells leading to the generation of TICs. 7. Importance of TGFb-induced EMT in metastasis Since metastasis is the foremost cause of cancer lethality, a deeper understanding of the mechanisms that promote metastasis is needed [95]. TGFb signaling has been linked to the process of metastasis in various cancer types, including breast and lung [95,96]. Current evidence supports the notion that EMT facilitates invasiveness and intravasation into lymph or blood vessels of cancer cells as one of the early manifestations of metastasis [5]. In addition to the direct effects of EMT on cancer cells, invasive tumor cells that have undergone EMT secrete many growth factors and chemokines that stimulate and recruit stromal cells, facilitating migration and intravasation of the tumor cell [95,97]. From a TGFb perspective, it is important to elucidate whether TGFb, in addition to the induction of EMT that initiates the cascade towards metastasis, continues to have active roles during additional steps in the trajectory that guides tumor cells to form metastases. 7.1. The dual role of TGFb in cancer progression and its links to EMT as established in mouse cancer models A suppressive role of TGFb1 on breast epithelial growth in vivo emerged from early studies of transgenic mice expressing an active, mature TGFb1 transgene in the mammary gland [98]. Under

C.-H. Heldin et al. / FEBS Letters 586 (2012) 1959–1970

normal growth conditions, the conditional transgene suppressed partially the development of the female mammary duct due to hypoproliferation of epithelial cells. When the TGFb1 mouse was crossed with an oncogenic mouse expressing the EGF family receptor Neu/Erb2 in the breast, the primary breast tumors developed equally well as the tumors induced by the Neu oncogene alone, demonstrating that the oncogene acted dominantly over the antiproliferative TGFb1 [99]. Surprisingly, the Neu/TGFb1 mice showed increased numbers of circulating tumor cells and lung metastases, while the tumor cells expressed protein markers that supported the presence of EMT and enhanced invasive capacity, corroborating the established model that EGF/Ras and TGFb signaling synergistically enhance EMT. The pro-metastatic effect of TGFb1 in breast cancer was further demonstrated by an inducible TGFb1 transgene expressed together with the polyomavirus middle T antigen that primed oncogenesis in the breast, leading to accelerated lung metastasis in a doxycycline-dependent manner [100]. Parallel studies using TGFb receptor transgenes provided additional and mechanistically detailed evidence for the dual response of breast cancer to TGFb signaling [101]. A constitutively active point mutant type I TGFb receptor expressed in the breast together with the Neu oncogene delayed the onset of primary breast cancer, but eventually, when the tumors grew, it enhanced the rate of lung metastasis. In contrast, a dominant negative, deletion mutant type II TGFb receptor accelerated the onset of primary tumor growth, but also reduced the rate of lung metastasis [101]. Knockout of the TGFb type II receptor in the mammary epithelium led to hyperplasia in the mammary gland of the normal mouse consistent with a growth inhibitory action of TGFb in normal epithelial biogenesis [102]. However, crossing of the TGFb type II deficient mouse with a mouse expressing the oncogenic polyoma middle T antigen in its breast, caused a more rapid onset of primary tumor growth, but, in contrast to the findings of Siegel and colleagues [101], it increased lung metastases [102]. The reason for the apparent discrepancy between the two studies [101,102] is not known. The above studies in the breast are supported by analogous studies in the skin. Targeted expression of TGFb1 in keratinocytes followed by a chemical carcinogenesis protocol, showed a reduced number of benign primary skin lesions, but eventually, as carcinogenesis progressed, TGFb1 induced a higher rate of malignant tumors with spindle-like carcinoma cells, thus providing one of the first demonstrations of TGFb-induced EMT in vivo [103]. The same result was reached by an inducible TGFb1 transgene in the skin and chemical carcinogenesis, showing that when TGFb1 was induced early it could suppress tumor growth, whereas when TGFb1 was induced at the papilloma stage, it actually promoted invasive tumor growth and metastasis [104]. The metastatic spindle cells exhibited strong signs of EMT in this model as well. Interestingly, co-expression of the TGFb1 transgene and the dominant negative TGFb type II receptor mutant in keratinocytes resulted in loss of the EMT response, as judged by epithelial and mesenchymal protein marker analysis in the growing tumors; however, the primary tumor cells that expressed epithelial proteins (e.g. E-cadherin) were able to invade and metastasize with an even higher rate [105]. The loss of EMT was correlated with decreased activation of Smad signaling, whereas the enhanced invasiveness and metastasis correlated with enhanced non-Smad (Rho/Rac GTPase and Erk MAPK) signaling. Transgenic mice expressing the dominant negative mutant type II receptor of TGFb in basal and follicular skin cells showed unperturbed normal tissue homeostasis [106]. Upon chemical carcinogenic challenge of these mice, the skin cells hyperproliferated and developed a higher number of faster growing carcinomas, supporting a tumor suppressor action of TGFb in the skin. In a parallel model, combination of oncogenic K- or H-Ras expression with knockout of the TGFb type II receptor in epithelial skin cells of

1965

the head and neck, led to dramatic tumor growth and metastasis, associated with enhanced endogenous TGFb1 production [107]. The reason why loss of the TGFb type II receptor from skin cells does not perturb normal tissue homeostasis is due to a compensation of the enhanced hyperproliferation of the skin cells by an enhanced apoptosis [108]. In the same skin type II receptor knockout model, spontaneous squamous carcinomas were observed in the rectum and genitals since the epithelial cells of these organs proliferate constantly during adult life. When the loss of the type II receptor was combined with an oncogenic Ras mutation, tumorigenesis was accelerated and the affected keratinocytes exhibited enhanced invasiveness [108]. The skin studies in mice nicely recapitulate the findings from human cancer patients with skin hyperplasias and head-and-neck cancer [107,108]. 7.2. The tumor cell environment is important for TGFb promotion of tumor progression The above example of skin carcinogenesis suggests that cancer cell autonomous TGFb signaling is responsible for the EMT; however, carcinoma-derived TGFb may act also on other cell types in the tumor environment to elicit invasiveness and metastasis. This has indeed been firmly proven in breast, skin, prostate and stomach cancer mouse models. Fibroblast-specific knockout of the TGFb type II receptor led to increased numbers of stromal cells and enhanced hyperplasia of the adjacent epithelium in the prostate and stomach [109], possibly by increased production of HGF by the fibroblasts, acting on the epithelial cells. It is noteworthy, that HGF, also known as scatter factor, was the first growth factor linked with the process of EMT [110]. According to this mouse model, the tumor suppressor action of TGFb is elicited by signaling in tumor stromal fibroblasts and not by signaling in the carcinoma cells. A similar action of TGFb was observed in breast cancer, by xenografting breast cancer cells into the same fibroblast-specific TGFb type II receptor knockout mouse [111]. Primary breast cancer growth and lung metastasis were enhanced, and correlated with overproduction of HGF and its receptor c-Met and enhanced signaling via Stat3 and Erk MAPK in the carcinoma cells. In addition, the enhanced lung metastasis in mice with conditional deletion of the TGFb type II receptor in breast cancers driven by the polyoma middle T antigen, as described above, was shown to be facilitated by myeloid cells of the Gr-1+CD11b+ phenotype [112]. The myeloid cells infiltrated the invasive front of the breast carcinoma by responding to the chemokines SDF-1 and CXCL5 via their corresponding receptors CXCR4 and CXCR2, and led to enhanced production of TGFb1 and metalloproteinase activity that promoted invasiveness and metastasis. In this breast cancer model, the primary tumor stroma underwent significant alterations that correlated with increased primary tumor size; the importance of a tumor suppressor action of TGFb, acting at the level of tumor initiation, progression and metastasis, was confirmed once again [113]. The link between tumor-derived TGFb and various chemokines of the CXCL family that mediate the communication between carcinoma cells and their stromal neighbors in breast cancer has also been verified in human cancer samples using mRNA expression and statistical correlation analyses [114]. 7.3. The pro-metastatic actions of TGFb include EMT but also other processes critical for metastasis The central link between TGFb signaling, EMT and metastasis was pioneered by studies using the mouse breast cancer cell model EpRas that overexpressed the H-Ras oncogene [115,116]. Engineering the dominant negative TGFb type II receptor mutant into these cells was shown to block normal cell-autonomous TGFb signaling, inhibit EMT, delay primary tumor growth in xenografted

1966

C.-H. Heldin et al. / FEBS Letters 586 (2012) 1959–1970

immunocompromised mice, and decrease lung metastasis [116]. These studies established a central role of TGFb in every step of cancer progression and a strong correlation between EMT and metastatic potential. In addition, the EpRas model demonstrated the need of crosstalk between TGFb signaling and Ras-mediated Erk/ MAPK and PI3-kinase signaling in order to elicit both EMT and metastasis [116,117]. In this model, TGFb induces the secretion of the cytokine ILEI, whose action is critical for metastasis [118]. In agreement with the necessity of crosstalk between TGFb and Ras signaling, ILEI leads to downstream signaling mediated by secretion of PDGF and consecutive Stat3 activation, which function independent of the TGFb pathway to drive EMT and metastasis in a model of hepatocellular carcinoma [86]. Thus, in these models, TGFb acts by inducing ILEI and then ILEI takes over by inducing PDGF signaling towards EMT and metastasis. Similar to the EpRas cell model, hepatocarcinoma cells were shown to undergo EMT in response to TGFb by inducing expression of the CXCR4 receptor, which endowed them with responsiveness to SDF-1a, thus promoting survival necessary for efficient invasiveness [119]. EMT has also been described in studies of invasive metastatic myeloma, where it was shown to be driven by the concerted action of multiple growth factors, including TGFb, EGF and HGF [120]. These growth factors are secreted by myeloid-derived suppressor cells that infiltrate the growing tumor by responding to chemotactic recruitment mediated by the chemokine CXCL5 that is secreted by the myeloma cells and acting on the suppressor cells found within the bone marrow. In addition to the sequential waves of growth factor, cytokine and chemokine secretion induced by TGFb, additional regulators are involved in metastasis. Some of these regulators control both EMT and invasiveness, whereas others affect only late stages of metastasis, and possibly act only at the distant site of secondary tumor growth during colonization. In invasive breast cancers of the basal-like subtype, high expression of annexin A1 has been recorded [121]. Annexin A1 acts as a positive regulator of TGFb signaling and promotes EMT and invasiveness; its downregulation using RNAi blocks lung metastases. A mechanistically interesting link between TGFb-driven EMT and the onset of metastasis has been illustrated by studies of breast cancer cells growing in three-dimensional organotypic cultures and the regulation of E-cadherin expression in response to TGFb [122]. Downregulation of E-cadherin prepares cancer cells for invasiveness by allowing a switch in the expression pattern of specific integrin receptors; induction of integrin b1 plays a critical role in invasiveness and metastasis. Studies of breast cancer cell invasion in host mice using intravital imaging technology permitted the observation of two types of tumor cell motility [123]. Single-cell motility was shown to be driven by active response to TGFb found in the tumor microenvironment, involvement of several signaling mediators such as Smad4 and Rho GTPases, and directional movement towards blood vessels. While this type of migration was fully blocked by inhibitors of TGFb receptor kinases, the tumor cells continued their invasive behavior even in the presence of the TGFb inhibitor, based on the second type of motility that involved cohorts of cells that migrated primarily towards the lymphatic vessels instead of blood vessels. Based on this study, metastatic dissemination of breast cancer to the lung requires transient TGFb signaling, and involves solitary invasive cells that presumably have undergone EMT, which metastasize using blood but not lymphatic vessels [124]. An important target of TGFb signaling that participates in lung metastasis of breast cancer cells is the angiopoietin-like 4, a direct gene target of Smad signaling [125]. While TGFb induces angiopoietin-like 4 expression at the primary tumor prior to metastasis, the functional role of this secreted protein is elicited only later during metastatic colonization, as the angiopoietin-like 4 protein assists during the process of extravasation by opening the strongly adhering endo-

thelial cells in capillary vessels of the lung, thus facilitating the deposition of tumor cells to the metastatic site. In contrast to the studies on breast cancer metastasis to lung that emphasizes transient TGFb signaling, additional studies of breast cancer metastasis to bone concluded that TGFb signaling is important both during the early phase of metastasis and during the colonization in the bone, where TGFb derived from the bone storages after the destructive action of osteoclasts could assist in the survival of the metastatic cells in the new niche [126,127]. These findings using modern non-invasive imaging technology confirm earlier pioneering studies of breast cancer metastasis to bone, which identified that TGFb acts at the osteolytic metastatic site, where it induces expression of the cytokine interleukin-11 (IL-11) and the pro-angiogenic growth factor connective tissue growth factor (CTGF), affecting the action of osteoclasts and angiogenesis at the new metastatic lesion [128]. Consequently, abrogation of TGFb signaling in the primary breast cancer cells by means of silencing the Smad4 gene, prior to metastatic spread in the host mouse, was detrimental to the metastatic colonization and inhibited the critical responses of IL-11 and CTGF [27,129]. These genetic experiments highlight the possible relevance of using TGFb inhibitors to prevent EMT and cancer metastasis. 7.4. Can TGFb inhibitors be used as anti-metastatic agents with clinical efficacy? The strong evidence for pro-tumorigenic and pro-metastatic actions of TGFb, suggests that inhibitors of this signaling pathway may be useful to treat advanced cancer [130,131]. Whereas the tumor suppressor action of TGFb complicates the use of TGFb inhibitors for cancer treatment, pre-clinical studies using anti-TGFb inhibitors show some promising results. The Fc:TGFb type II receptor fusion protein acts as a decoy for all isoforms of TGFb; administration of this TGFb inhibitor intravenously in the polyoma middle T antigen transgenic breast cancer model, had no impact on primary tumor growth but effectively reduced breast cancer invasiveness and lung metastases [132]. Similar anti-metastatic efficacy was found when this inhibitor was used in syngeneic xenograft breast cancer models of 4T1 and EMT-6 cells that metastasize to lungs [132]. Moreover, expression of the decoy protein as a transgene in the mammary gland of mice [133] and crossing with the Neu breast oncomouse or injecting isogenic melanoma cells in the tail vein, effectively reduced lung metastases in all cases. Similar results were obtained by a neutralizing pan-TGFb antibody when administered systemically to the polyoma middle T antigen breast cancer model followed by ionizing radiation or doxorubicin therapy [134]. The anti-TGFb antibody blocked lung metastasis and normalized the adverse effects of ionizing radiation and doxorubicin, as both of these therapeutic modalities enhanced secretion of TGFb by tumor cells. A second study with the anti-TGFb monoclonal antibody 1D11 and the xenograft metastasis model of 4T1 breast cancer cells, demonstrated strong efficacy against lung metastasis [135]. In addition to direct inhibition of the TGFb ligand, interference with factors that assist TGFb signaling specifically during cancer metastasis may be an alternative and more targeted method of therapy. Two recent studies highlight this possibility. Studies of EMT and metastasis by lung cancer cells in response to TGFb uncovered a role of the transcription factor peroxisome proliferator-activated receptor c (PPARc) as an inhibitor of Smad3 transcriptional function during EMT and metastasis [136]. Natural ligands of the PPARc protein such as troglitazone and rosiglitazone activate PPARc, which then blocks Smad3 function and thus effectively inhibit EMT in vitro and metastasis in xenografted mice in vivo. Moreover, the transmembrane protein Klotho that is mainly expressed by kidney cells can be shed, and this secreted form binds

C.-H. Heldin et al. / FEBS Letters 586 (2012) 1959–1970

to the TGFb type II receptor and competitively inhibits its binding to the natural TGFb ligand [137]. Klotho can therefore act as a TGFb inhibitor and administration of Klotho was effective against metastasis in a xenograft mouse model [137]. Klotho does not show specificity for the TGFb receptor and is also known to block Wnt and IGF signaling. Finally, low molecular weight inhibitors that inactivate the TGFb receptor kinases have been developed [130]. A dual type II and type I receptor kinase inhibitor, LY2109761, gave impressive inhibition of metastasis in an orthotopic pancreatic cancer model [138]. A comparative study that examined breast cancer metastasis to lung or bone after treatment with the anti-TGFb 1D11 antibody and the receptor kinase inhibitor LY2109761 concluded that both agents were equally efficient to inhibit metastasis [139]. The anti-TGFb therapy was equally effective in lung or bone metastasis and the inhibitor blocked both metastatic angiogenesis in the lung and osteoclastic activity in the bone lesions. A recent study has examined the well-established phenomenon of acquired drug resistance to inhibitors of signaling molecules in cancer, now focusing on the TGFb receptor inhibitor LY2109761 [140]. Effects of chronic administration of the inhibitor in mice with skin cancer caused by tumor cell transplantation or chemical carcinogenesis were analyzed; the study revealed for the first time development of drug-resistant tumors with signs of EMT and aggressive invasiveness. Thus, the use of the new generation anti-TGFb signaling agents progresses with caution and the results of the ongoing clinical trials by Eli Lilly are awaited with great anticipation. It is possible that such anti-TGFb drugs may be used against metastasis only for certain tumors, in limited doses, and in combination with other inhibitors that target more selectively EMT or the following metastatic cascade. 8. Future perspectives The role of TGFb in EMT, carcinoma invasiveness and metastasis to multiple organs has been firmly established based on in vitro and in vivo studies. Remaining open challenges are the elucidation of the comprehensive network of molecules that initiate and execute the EMT program. Identification of key enzymes within this network will be critical from a therapeutic perspective. Furthermore, the link between mechanisms that elicit TGFb-induced EMT and mechanisms that maintain cancer stem cells in different organs is just emerging and awaits long series of investigations. Finally, the precise role of EMT in the multi-stage metastatic process of human cancer needs to be clarified. Such studies will involve analyses of tumor tissues and blood samples with circulating tumor cells with regard to expression and activation of the networks of proteins that participate in the EMT program. Acknowledgments We acknowledge funding by the Ludwig Institute for Cancer Research, the Swedish Cancer Society, the Swedish Research Council and the Network of Excellence ‘‘ENFIN’’ under the European Union FP6 program. We thank all past and present members of the TGFb signaling group for their contributions to the scientific work emanating from our laboratory. Special thanks go to Ingegärd Schiller for assistance with the preparation of the manuscript. We apologize to those authors whose relevant work has not been included in this review article due to space limitations. We declare no conflict of interest. References [1] Moustakas, A. and Heldin, C.-H. (2009) The regulation of TGFb signal transduction. Development 136, 3699–3714.

1967

[2] Massagué, J. (2008) TGFb in cancer. Cell 134, 215–230. [3] Gordon, K.J. and Blobe, G.C. (2008) Role of transforming growth factor-b superfamily signaling pathways in human disease. Biochim. Biophys. Acta 1782, 197–228. [4] Levy, L. and Hill, C.S. (2006) Alterations in components of the TGF-b superfamily signaling pathways in human cancer. Cytokine Growth Factor Rev. 17, 41–58. [5] Thiery, J.P., Acloque, H., Huang, R.Y. and Nieto, M.A. (2009) Epithelialmesenchymal transitions in development and disease. Cell 139, 871–890. [6] Kang, J.S., Liu, C. and Derynck, R. (2009) New regulatory mechanisms of TGF-b receptor function. Trends Cell Biol. 19, 385–394. [7] Sorrentino, A., Thakur, N., Grimsby, S., Marcusson, A., von Bulow, V., Schuster, N., Zhang, S., Heldin, C.-H. and Landström, M. (2008) The type I TGF-b receptor engages TRAF6 to activate TAK1 in a receptor kinase-independent manner. Nat. Cell Biol. 10, 1199–1207. [8] Yamashita, M., Fatyol, K., Jin, C., Wang, X., Liu, Z. and Zhang, Y.E. (2008) TRAF6 mediates Smad-independent activation of JNK and p38 by TGF-b. Mol. Cell 31, 918–924. [9] Mu, Y., Sundar, R., Thakur, N., Ekman, M., Gudey, S.K., Yakymovych, M., Hermansson, A., Dimitriou, H., Bengoechea-Alonso, M.T., Ericsson, J., Heldin, C.-H. and Landström, M. (2011) TRAF6 ubiquitinates TGFb type I receptor to promote its cleavage and nuclear translocation in cancer. Nat. Commun. 2, 330. [10] Miettinen, P.J., Ebner, R., Lopez, A.R. and Derynck, R. (1994) TGF-b induced transdifferentiation of mammary epithelial cells to mesenchymal cells: Involvement of type I receptors. J. Cell Biol. 127, 2021–2036. [11] López-Novoa, J.M. and Nieto, M.A. (2009) Inflammation and EMT: an alliance towards organ fibrosis and cancer progression. EMBO Mol. Med. 1, 303–314. [12] Iwano, M., Plieth, D., Danoff, T.M., Xue, C., Okada, H. and Neilson, E.G. (2002) Evidence that fibroblasts derive from epithelium during tissue fibrosis. J. Clin. Invest. 110, 341–350. [13] Zeisberg, M., Yang, C., Martino, M., Duncan, M.B., Rieder, F., Tanjore, H. and Kalluri, R. (2007) Fibroblasts derive from hepatocytes in liver fibrosis via epithelial to mesenchymal transition. J. Biol. Chem. 282, 23337–23347. [14] Kim, K.K., Kugler, M.C., Wolters, P.J., Robillard, L., Galvez, M.G., Brumwell, A.N., Sheppard, D. and Chapman, H.A. (2006) Alveolar epithelial cell mesenchymal transition develops in vivo during pulmonary fibrosis and is regulated by the extracellular matrix. Proc. Natl. Acad. Sci. USA 103, 13180– 13185. [15] Zeisberg, E.M., Tarnavski, O., Zeisberg, M., Dorfman, A.L., McMullen, J.R., Gustafsson, E., Chandraker, A., Yuan, X., Pu, W.T., Roberts, A.B., Neilson, E.G., Sayegh, M.H., Izumo, S. and Kalluri, R. (2007) Endothelial-to-mesenchymal transition contributes to cardiac fibrosis. Nat. Med. 13, 952–961. [16] Zeisberg, E.M., Potenta, S.E., Sugimoto, H., Zeisberg, M. and Kalluri, R. (2008) Fibroblasts in kidney fibrosis emerge via endothelial-to-mesenchymal transition. J. Am. Soc. Nephrol. 19, 2282–2287. [17] Schmalhofer, O., Brabletz, S. and Brabletz, T. (2009) E-cadherin, beta-catenin, and ZEB1 in malignant progression of cancer. Cancer Metastasis Rev. 28, 151– 166. [18] Wicki, A., Lehembre, F., Wick, N., Hantusch, B., Kerjaschki, D. and Christofori, G. (2006) Tumor invasion in the absence of epithelial-mesenchymal transition: podoplanin-mediated remodeling of the actin cytoskeleton. Cancer Cell 9, 261–272. [19] Peinado, H., Olmeda, D. and Cano, A. (2007) Snail, Zeb and bHLH factors in tumour progression: an alliance against the epithelial phenotype? Nat. Rev. Cancer 7, 415–428. [20] Mani, S.A., Yang, J., Brooks, M., Schwaninger, G., Zhou, A., Miura, N., Kutok, J.L., Hartwell, K., Richardson, A.L. and Weinberg, R.A. (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. [21] Thuault, S., Valcourt, U., Petersen, M., Manfioletti, G., Heldin, C.-H. and Moustakas, A. (2006) Transforming growth factor-b employs HMGA2 to elicit epithelial–mesenchymal transition. J. Cell Biol. 174, 175–183. [22] Thuault, S., Tan, E.-J., Peinado, H., Cano, A., Heldin, C.-H. and Moustakas, A. (2008) HMGA2 and Smads coregulate SNAIL1 expression during induction of epithelial-to-mesenchymal transition. J. Biol. Chem. 283, 33437–33446. [23] Smit, M.A., Geiger, T.R., Song, J.Y., Gitelman, I. and Peeper, D.S. (2009) A TwistSnail axis critical for TrkB-induced epithelial-mesenchymal transition-like transformation, anoikis resistance, and metastasis. Mol. Cell. Biol. 29, 3722– 3737. [24] Dave, N., Guaita-Esteruelas, S., Gutarra, S., Frias, A., Beltran, M., Peiro, S. and de Herreros, A.G. (2011) Functional cooperation between Snail1 and twist in the regulation of ZEB1 expression during epithelial to mesenchymal transition. J. Biol. Chem. 286, 12024–12032. [25] Guaita, S., Puig, I., Franci, C., Garrido, M., Dominguez, D., Batlle, E., Sancho, E., Dedhar, S., de Herreros, A.G. and Baulida, J. (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. [26] Casas, E., Kim, J., Bendesky, A., Ohno-Machado, L., Wolfe, C.J. and Yang, J. (2011) Snail2 is an essential mediator of Twist1-induced epithelial mesenchymal transition and metastasis. Cancer Res. 71, 245–254. [27] Deckers, M., van Dinther, M., Buijs, J., Que, I., Lowik, C., van der Pluijm, G. and ten Dijke, P. (2006) The tumor suppressor Smad4 is required for transforming growth factor b-induced epithelial to mesenchymal transition and bone metastasis of breast cancer cells. Cancer Res. 66, 2202–2209.

1968

C.-H. Heldin et al. / FEBS Letters 586 (2012) 1959–1970

[28] Valcourt, U., Kowanetz, M., Niimi, H., Heldin, C.-H. and Moustakas, A. (2005) TGF-b and the Smad signaling pathway support transcriptomic reprogramming during epithelial mesenchymal cell transition. Mol. Biol. Cell 16, 1987–2002. [29] Roberts, A.B., Tian, F., Byfield, S.D., Stuelten, C., Ooshima, A., Saika, S. and Flanders, K.C. (2006) Smad3 is key to TGF-b-mediated epithelial-tomesenchymal transition, fibrosis, tumor suppression and metastasis. Cytokine Growth Factor Rev. 17, 19–27. [30] Piek, E., Moustakas, A., Kurisaki, A., Heldin, C.-H. and ten Dijke, P. (1999) TGFb type I receptor/ALK5 and SMAD proteins mediate epithelial to mesenchymal transdifferentiation in NMuMG breast epithelial cells. J. Cell Sci. 112, 4557–4568. [31] Hoot, K.E., Lighthall, J., Han, G., Lu, S.L., Li, A., Ju, W., Kulesz-Martin, M., Bottinger, E. and Wang, X.J. (2008) Keratinocyte-specific Smad2 ablation results in increased epithelial-mesenchymal transition during skin cancer formation and progression. J. Clin. Invest. 118, 2722–2732. [32] Runyan, C.E., Hayashida, T., Hubchak, S., Curley, J.F. and Schnaper, H.W. (2009) Role of SARA (SMAD anchor for receptor activation) in maintenance of epithelial cell phenotype. J. Biol. Chem. 284, 25181–25189. [33] Soond, S.M. and Chantry, A. (2011) Selective targeting of activating and inhibitory Smads by distinct WWP2 ubiquitin ligase isoforms differentially modulates TGFb signalling and EMT. Oncogene 30, 2451–2462. [34] Dupont, S., Mamidi, A., Cordenonsi, M., Montagner, M., Zacchigna, L., Adorno, M., Martello, G., Stinchfield, M.J., Soligo, S., Morsut, L., Inui, M., Moro, S., Modena, N., Argenton, F., Newfeld, S.J. and Piccolo, S. (2009) FAM/USP9x, a deubiquitinating enzyme essential for TGFb signaling, controls Smad4 monoubiquitination. Cell 136, 123–135. [35] He, W., Dorn, D.C., Erdjument-Bromage, H., Tempst, P., Moore, M.A. and Massagué, J. (2006) Hematopoiesis controlled by distinct TIF1c and Smad4 branches of the TGFb pathway. Cell 125, 929–941. [36] Hesling, C., Fattet, L., Teyre, G., Jury, D., Gonzalo, P., Lopez, J., Vanbelle, C., Morel, A.P., Gillet, G., Mikaelian, I. and Rimokh, R. (2011) Antagonistic regulation of EMT by TIF1c and Smad4 in mammary epithelial cells. EMBO Rep. 12, 665–672. [37] Tan, E.-J., Thuault, S., Caja, L., Carletti, T., Heldin, C.-H., and Moustakas, A. (in press). Regulation of transcription factor Twist expression by the DNA architectural protein high mobility group A2 during epithelial-tomesenchymal transition. J. Biol. Chem. [38] Peinado, H., Quintanilla, M. and Cano, A. (2003) Transforming growth factor b-1 induces snail transcription factor in epithelial cell lines: mechanisms for epithelial mesenchymal transitions. J. Biol. Chem. 278, 21113–21123. [39] Yin, X., Wolford, C.C., Chang, Y.S., McConoughey, S.J., Ramsey, S.A., Aderem, A. and Hai, T. (2010) ATF3, an adaptive-response gene, enhances TGFb signaling and cancer-initiating cell features in breast cancer cells. J. Cell Sci. 123, 3558– 3565. [40] Huber, M.A., Azoitei, N., Baumann, B., Grunert, S., Sommer, A., Pehamberger, H., Kraut, N., Beug, H. and Wirth, T. (2004) NF-jB is essential for epithelialmesenchymal transition and metastasis in a model of breast cancer progression. J. Clin. Invest. 114, 569–581. [41] Papageorgis, P., Lambert, A.W., Ozturk, S., Gao, F., Pan, H., Manne, U., Alekseyev, Y.O., Thiagalingam, A., Abdolmaleky, H.M., Lenburg, M. and Thiagalingam, S. (2010) Smad signaling is required to maintain epigenetic silencing during breast cancer progression. Cancer Res. 70, 968–978. [42] McDonald, O.G., Wu, H., Timp, W., Doi, A. and Feinberg, A.P. (2011) Genomescale epigenetic reprogramming during epithelial-to-mesenchymal transition. Nat. Struct. Mol. Biol. 18, 867–874. [43] Wu, M.Z., Tsai, Y.P., Yang, M.H., Huang, C.H., Chang, S.Y., Chang, C.C., Teng, S.C. and Wu, K.J. (2011) Interplay between HDAC3 and WDR5 is essential for hypoxia-induced epithelial-mesenchymal transition. Mol. Cell 43, 811–822. [44] Shan, B., Yao, T.P., Nguyen, H.T., Zhuo, Y., Levy, D.R., Klingsberg, R.C., Tao, H., Palmer, M.L., Holder, K.N. and Lasky, J.A. (2008) Requirement of HDAC6 for transforming growth factor-b1-induced epithelial-mesenchymal transition. J. Biol. Chem. 283, 21065–21073. [45] Wang, X., Nie, J., Zhou, Q., Liu, W., Zhu, F., Chen, W., Mao, H., Luo, N., Dong, X. and Yu, X. (2008) Downregulation of Par-3 expression and disruption of Par complex integrity by TGF-b during the process of epithelial to mesenchymal transition in rat proximal epithelial cells. Biochim. Biophys. Acta 1782, 51– 59. [46] Ozdamar, B., Bose, R., Barrios-Rodiles, M., Wang, H.R., Zhang, Y. and Wrana, J.L. (2005) Regulation of the polarity protein Par6 by TGFb receptors controls epithelial cell plasticity. Science 307, 1603–1609. [47] Spaderna, S., Schmalhofer, O., Wahlbuhl, M., Dimmler, A., Bauer, K., Sultan, A., Hlubek, F., Jung, A., Strand, D., Eger, A., Kirchner, T., Behrens, J. and Brabletz, T. (2008) The transcriptional repressor ZEB1 promotes metastasis and loss of cell polarity in cancer. Cancer Res. 68, 537–544. [48] Burk, U., Schubert, J., Wellner, U., Schmalhofer, O., Vincan, E., Spaderna, S. and Brabletz, T. (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. [49] Gregory, P.A., Bert, A.G., Paterson, E.L., Barry, S.C., Tsykin, A., Farshid, G., Vadas, M.A., Khew-Goodall, Y. and Goodall, G.J. (2008) The miR-200 family and miR-205 regulate epithelial to mesenchymal transition by targeting ZEB1 and SIP1. Nat. Cell Biol. 10, 593–601. [50] Korpal, M., Lee, E.S., Hu, G. and Kang, Y. (2008) The miR-200 family inhibits epithelial-mesenchymal transition and cancer cell migration by direct

[51]

[52]

[53]

[54]

[55]

[56]

[57]

[58] [59]

[60]

[61]

[62]

[63]

[64]

[65]

[66]

[67]

[68]

[69]

[70]

[71]

[72]

[73]

targeting of E-cadherin transcriptional repressors ZEB1 and ZEB2. J. Biol. Chem. 283, 14910–14914. Park, S.M., Gaur, A.B., Lengyel, E. and Peter, M.E. (2008) The miR-200 family determines the epithelial phenotype of cancer cells by targeting the Ecadherin repressors ZEB1 and ZEB2. Genes Dev. 22, 894–907. Iliopoulos, D., Polytarchou, C., Hatziapostolou, M., Kottakis, F., Maroulakou, I.G., Struhl, K., and Tsichlis, P.N. (2009). MicroRNAs differentially regulated by Akt isoforms control EMT and stem cell renewal in cancer cells. Sci. Signal. 2, ra62. Beltran, M., Puig, I., Pena, C., Garcia, J.M., Alvarez, A.B., Pena, R., Bonilla, F. and de Herreros, A.G. (2008) A natural antisense transcript regulates Zeb2/Sip1 gene expression during Snail1-induced epithelial-mesenchymal transition. Genes Dev. 22, 756–769. Gregory, P.A., Bracken, C.P., Smith, E., Bert, A.G., Wright, J.A., Roslan, S., Morris, M., Wyatt, L., Farshid, G., Lim, Y.Y., Lindeman, G.J., Shannon, M.F., Drew, P.A., Khew-Goodall, Y. and Goodall, G.J. (2011) An autocrine TGF-b/ZEB/miR-200 signaling network regulates establishment and maintenance of epithelialmesenchymal transition. Mol. Biol. Cell 22, 1686–1698. Eades, G., Yao, Y., Yang, M., Zhang, Y., Chumsri, S. and Zhou, Q. (2011) MiR200a regulates SIRT1 expression and epithelial to mesenchymal transition (EMT)-like transformation in mammary epithelial cells. J. Biol. Chem. 286, 25992–26002. Kong, W., Yang, H., He, L., Zhao, J.J., Coppola, D., Dalton, W.S. and Cheng, J.Q. (2008) MicroRNA-155 is regulated by the transforming growth factor b/Smad pathway and contributes to epithelial cell plasticity by targeting RhoA. Mol. Cell. Biol. 28, 6773–6784. Zhou, Q., Fan, J., Ding, X., Peng, W., Yu, X., Chen, Y. and Nie, J. (2010) TGF-binduced MiR-491-5p expression promotes Par-3 degradation in rat proximal tubular epithelial cells. J. Biol. Chem. 285, 40019–40027. Ma, L., Teruya-Feldstein, J. and Weinberg, R.A. (2007) Tumour invasion and metastasis initiated by microRNA-10b in breast cancer. Nature 449, 682–688. Shirakihara, T., Horiguchi, K., Miyazawa, K., Ehata, S., Shibata, T., Morita, I., Miyazono, K. and Saitoh, M. (2011) TGF-b regulates isoform switching of FGF receptors and epithelial-mesenchymal transition. EMBO J. 30, 783–795. Horiguchi, K., Sakamoto, K., Koinuma, D., Semba, K., Inoue, A., Inoue, S., Fujii, H., Yamaguchi, A., Miyazawa, K., Miyazono, K., and Saitoh, M. (in press). TGF-b drives epithelial-mesenchymal transition through dEF1-mediated downregulation of ESRP. Oncogene. Warzecha, C.C., Sato, T.K., Nabet, B., Hogenesch, J.B. and Carstens, R.P. (2009) ESRP1 and ESRP2 are epithelial cell-type-specific regulators of FGFR2 splicing. Mol. Cell 33, 591–601. Warzecha, C.C., Shen, S., Xing, Y. and Carstens, R.P. (2009) The epithelial splicing factors ESRP1 and ESRP2 positively and negatively regulate diverse types of alternative splicing events. RNA Biol. 6, 546–562. Chaudhury, A., Hussey, G.S., Ray, P.S., Jin, G., Fox, P.L. and Howe, P.H. (2010) TGF-b-mediated phosphorylation of hnRNP E1 induces EMT via transcriptselective translational induction of Dab2 and ILEI. Nat. Cell Biol. 12, 286–293. Hussey, G.S., Chaudhury, A., Dawson, A.E., Lindner, D.J., Knudsen, C.R., Wilce, M.C., Merrick, W.C. and Howe, P.H. (2011) Identification of an mRNP complex regulating tumorigenesis at the translational elongation step. Mol. Cell 41, 419–431. Huber, M.A., Kraut, N. and Beug, H. (2005) Molecular requirements for epithelial-mesenchymal transition during tumor progression. Curr. Opin. Cell Biol. 17, 548–558. Yook, J.I., Li, X.Y., Ota, I., Hu, C., Kim, H.S., Kim, N.H., Cha, S.Y., Ryu, J.K., Choi, Y.J., Kim, J., Fearon, E.R. and Weiss, S.J. (2006) A Wnt-Axin2-GSK3b cascade regulates Snail1 activity in breast cancer cells. Nat. Cell Biol. 8, 1398–1406. Nawshad, A., Medici, D., Liu, C.C. and Hay, E.D. (2007) TGFb3 inhibits Ecadherin gene expression in palate medial-edge epithelial cells through a Smad2-Smad4-LEF1 transcription complex. J. Cell Sci. 120, 1646–1653. Vincent, T., Neve, E.P.A., Johnson, J.R., Kukalev, A., Rojo, F., Albanell, J., Pietras, K., Virtanen, I., Philipson, L., Leopold, P.L., Crystal, R.G., de Herreros, A.G., Moustakas, A., Pettersson, R.F. and Fuxe, J. (2009) A SNAIL1-SMAD3/4 transcriptional repressor complex promotes TGF-b mediated epithelialmesenchymal transition. Nat. Cell Biol. 11, 943–950. Kim, K.K., Wei, Y., Szekeres, C., Kugler, M.C., Wolters, P.J., Hill, M.L., Frank, J.A., Brumwell, A.N., Wheeler, S.E., Kreidberg, J.A. and Chapman, H.A. (2009) Epithelial cell a3b1 integrin links b-catenin and Smad signaling to promote myofibroblast formation and pulmonary fibrosis. J. Clin. Invest. 119, 213– 224. Zhou, B., Liu, Y., Kahn, M., Ann, D.K., Han, A., Wang, H., Nguyen, C., Flodby, P., Zhong, Q., Krishnaveni, M.S., Liebler, J.M., Minoo, P., Crandall, E.D., and Borok, Z. (in press). b-catenin/CBP-dependent regulation of TGF-b-mediated epithelial-mesenchymal transition (EMT) by smad3. J. Biol. Chem. Ripka, S., Konig, A., Buchholz, M., Wagner, M., Sipos, B., Kloppel, G., Downward, J., Gress, T. and Michl, P. (2007) WNT5A-target of CUTL1 and potent modulator of tumor cell migration and invasion in pancreatic cancer. Carcinogenesis 28, 1178–1187. Scheel, C., Eaton, E.N., Li, S.H., Chaffer, C.L., Reinhardt, F., Kah, K.J., Bell, G., Guo, W., Rubin, J., Richardson, A.L. and Weinberg, R.A. (2011) Paracrine and autocrine signals induce and maintain mesenchymal and stem cell states in the breast. Cell 145, 926–940. Moustakas, A. and Heldin, C.-H. (2007) Signaling networks guiding epithelial–mesenchymal transitions during embryogenesis and cancer progression. Cancer Sci. 98, 1512–1520.

C.-H. Heldin et al. / FEBS Letters 586 (2012) 1959–1970 [74] Timmerman, L.A., Grego-Bessa, J., Raya, A., Bertran, E., Perez-Pomares, J.M., Diez, J., Aranda, S., Palomo, S., McCormick, F., Izpisua-Belmonte, J.C. and de la Pompa, J.L. (2004) Notch promotes epithelial-mesenchymal transition during cardiac development and oncogenic transformation. Genes Dev. 18, 99–115. [75] Niessen, K., Fu, Y., Chang, L., Hoodless, P.A., McFadden, D. and Karsan, A. (2008) Slug is a direct Notch target required for initiation of cardiac cushion cellularization. J. Cell Biol. 182, 315–325. [76] Luna-Zurita, L., Prados, B., Grego-Bessa, J., Luxan, G., del Monte, G., Benguria, A., Adams, R.H., Perez-Pomares, J.M. and de la Pompa, J.L. (2010) Integration of a Notch-dependent mesenchymal gene program and Bmp2-driven cell invasiveness regulates murine cardiac valve formation. J. Clin. Invest. 120, 3493–3507. [77] Niimi, H., Pardali, K., Vanlandewijck, M., Heldin, C.-H. and Moustakas, A. (2007) Notch signaling is necessary for epithelial growth arrest by TGF-b. J. Cell Biol. 176, 695–707. [78] Nyhan, K.C., Faherty, N., Murray, G., Cooey, L.B., Godson, C., Crean, J.K. and Brazil, D.P. (2010) Jagged/Notch signalling is required for a subset of TGFb1 responses in human kidney epithelial cells. Biochim. Biophys. Acta 1803, 1386–1395. [79] Zavadil, J., Cermak, L., Soto-Nieves, N. and Bottinger, E.P. (2004) Integration of TGF-b/Smad and Jagged1/Notch signalling in epithelial-to-mesenchymal transition. EMBO J. 23, 1155–1165. [80] Sjölund, J., Boström, A.-K., Lindgren, D., Manna, S., Moustakas, A., Ljungberg, B., Johansson, M., Fredlund, E. and Axelson, H. (2011) The notch and TGF-b signaling pathways contribute to the aggressiveness of clear cell renal cell carcinoma. PLoS ONE 6, e23057. [81] Aoyagi-Ikeda, K., Maeno, T., Matsui, H., Ueno, M., Hara, K., Aoki, Y., Aoki, F., Shimizu, T., Doi, H., Kawai-Kowase, K., Iso, T., Suga, T., Arai, M. and Kurabayashi, M. (2011) Notch induces myofibroblast differentiation of alveolar epithelial cells via transforming growth factor-b-Smad3 pathway. Am. J. Respir. Cell Mol. Biol. 45, 136–144. [82] Ohashi, S., Natsuizaka, M., Naganuma, S., Kagawa, S., Kimura, S., Itoh, H., Kalman, R.A., Nakagawa, M., Darling, D.S., Basu, D., Gimotty, P.A., KleinSzanto, A.J., Diehl, J.A., Herlyn, M. and Nakagawa, H. (2011) A NOTCH3mediated squamous cell differentiation program limits expansion of EMTcompetent cells that express the ZEB transcription factors. Cancer Res. 71, 6836–6847. [83] Gotzmann, J., Fischer, A.N., Zojer, M., Mikula, M., Proell, V., Huber, H., Jechlinger, M., Waerner, T., Weith, A., Beug, H. and Mikulits, W. (2006) A crucial function of PDGF in TGF-b-mediated cancer progression of hepatocytes. Oncogene 25, 3170–3185. [84] Fischer, A.N., Fuchs, E., Mikula, M., Huber, H., Beug, H. and Mikulits, W. (2007) PDGF essentially links TGF-b signaling to nuclear b-catenin accumulation in hepatocellular carcinoma progression. Oncogene 26, 3395–3405. [85] van Zijl, F., Mair, M., Csiszar, A., Schneller, D., Zulehner, G., Huber, H., Eferl, R., Beug, H., Dolznig, H. and Mikulits, W. (2009) Hepatic tumor-stroma crosstalk guides epithelial to mesenchymal transition at the tumor edge. Oncogene 28, 4022–4033. [86] Lahsnig, C., Mikula, M., Petz, M., Zulehner, G., Schneller, D., van Zijl, F., Huber, H., Csiszar, A., Beug, H. and Mikulits, W. (2009) ILEI requires oncogenic Ras for the epithelial to mesenchymal transition of hepatocytes and liver carcinoma progression. Oncogene 28, 638–650. [87] Jechlinger, M., Sommer, A., Moriggl, R., Seither, P., Kraut, N., Capodiecci, P., Donovan, M., Cordon-Cardo, C., Beug, H. and Grunert, S. (2006) Autocrine PDGFR signaling promotes mammary cancer metastasis. J. Clin. Invest. 116, 1561–1570. [88] Hanahan, D. and Weinberg, R.A. (2011) Hallmarks of cancer: the next generation. Cell 144, 646–674. [89] Mani, S.A., Guo, W., Liao, M.J., Eaton, E.N., Ayyanan, A., Zhou, A.Y., Brooks, M., Reinhard, F., Zhang, C.C., Shipitsin, M., Campbell, L.L., Polyak, K., Brisken, C., Yang, J. and Weinberg, R.A. (2008) The epithelial-mesenchymal transition generates cells with properties of stem cells. Cell 133, 704–715. [90] Morel, A.P., Lievre, M., Thomas, C., Hinkal, G., Ansieau, S. and Puisieux, A. (2008) Generation of breast cancer stem cells through epithelialmesenchymal transition. PLoS ONE 3, e2888. [91] Battula, V.L., Evans, K.W., Hollier, B.G., Shi, Y., Marini, F.C., Ayyanan, A., Wang, R.Y., Brisken, C., Guerra, R., Andreeff, M. and Mani, S.A. (2010) Epithelialmesenchymal transition-derived cells exhibit multilineage differentiation potential similar to mesenchymal stem cells. Stem Cells 28, 1435–1445. [92] Caja, L., Bertran, E., Campbell, J., Fausto, N. and Fabregat, I. (2011) The transforming growth factor-b (TGF-b) mediates acquisition of a mesenchymal stem cell-like phenotype in human liver cells. J. Cell. Physiol. 226, 1214–1223. [93] Asiedu, M.K., Ingle, J.N., Behrens, M.D., Radisky, D.C. and Knutson, K.L. (2011) TGFb/TNF(a)-mediated epithelial-mesenchymal transition generates breast cancer stem cells with a claudin-low phenotype. Cancer Res. 71, 4707–4719. [94] Li, R., Liang, J., Ni, S., Zhou, T., Qing, X., Li, H., He, W., Chen, J., Li, F., Zhuang, Q., Qin, B., Xu, J., Li, W., Yang, J., Gan, Y., Qin, D., Feng, S., Song, H., Yang, D., Zhang, B., Zeng, L., Lai, L., Esteban, M.A. and Pei, D. (2010) A mesenchymal-toepithelial transition initiates and is required for the nuclear reprogramming of mouse fibroblasts. Cell Stem Cell 7, 51–63. [95] Nguyen, D.X., Bos, P.D. and Massagué, J. (2009) Metastasis: from dissemination to organ-specific colonization. Nat. Rev. Cancer 9, 274–284. [96] Bierie, B. and Moses, H.L. (2006) Tumour microenvironment: TGFb: the molecular Jekyll and Hyde of cancer. Nat. Rev. Cancer 6, 506–520.

1969

[97] Pietras, K. and Östman, A. (2010) Hallmarks of cancer: interactions with the tumor stroma. Exp. Cell Res. 316, 1324–1331. [98] Pierce Jr., D., Johnson, M.D., Matsui, Y., Robinson, S.D., Gold, L.I., Purchio, A.F., Daniel, C.W., Hogan, B. and Moses, H.L. (1993) Inhibition of mammary duct development but not alveolar outgrowth during pregnancy in transgenic mice expressing active TGF-b1. Genes Dev. 7, 2308–2317. [99] Muraoka, R.S., Koh, Y., Roebuck, L.R., Sanders, M.E., Brantley-Sieders, D., Gorska, A.E., Moses, H.L. and Arteaga, C.L. (2003) Increased malignancy of Neu-induced mammary tumors overexpressing active transforming growth factor b1. Mol. Cell. Biol. 23, 8691–8703. [100] Muraoka-Cook, R.S., Kurokawa, H., Koh, Y., Forbes, J.T., Roebuck, L.R., Barcellos-Hoff, M.H., Moody, S.E., Chodosh, L.A. and Arteaga, C.L. (2004) Conditional overexpression of active transforming growth factor b1 in vivo accelerates metastases of transgenic mammary tumors. Cancer Res. 64, 9002–9011. [101] Siegel, P.M., Shu, W., Cardiff, R.D., Muller, W.J. and Massagué, J. (2003) Transforming growth factor b signaling impairs Neu-induced mammary tumorigenesis while promoting pulmonary metastasis. Proc. Natl. Acad. Sci. USA 100, 8430–8435. [102] Forrester, E., Chytil, A., Bierie, B., Aakre, M., Gorska, A.E., Sharif-Afshar, A.R., Muller, W.J. and Moses, H.L. (2005) Effect of conditional knockout of the type II TGF-b receptor gene in mammary epithelia on mammary gland development and polyomavirus middle T antigen induced tumor formation and metastasis. Cancer Res. 65, 2296–2302. [103] Cui, W., Fowlis, D.J., Bryson, S., Duffie, E., Ireland, H., Balmain, A. and Akhurst, R.J. (1996) TGFb1 inhibits the formation of benign skin tumors, but enhances progression to invasive spindle carcinomas in transgenic mice. Cell 86, 531– 542. [104] Weeks, B.H., He, W., Olson, K.L. and Wang, X.J. (2001) Inducible expression of transforming growth factor b1 in papillomas causes rapid metastasis. Cancer Res. 61, 7435–7443. [105] Han, G., Lu, S.L., Li, A.G., He, W., Corless, C.L., Kulesz-Martin, M. and Wang, X.J. (2005) Distinct mechanisms of TGF-b1-mediated epithelial-to mesenchymal transition and metastasis during skin carcinogenesis. J. Clin. Invest. 115, 1714–1723. [106] Amendt, C., Schirmacher, P., Weber, H. and Blessing, M. (1998) Expression of a dominant negative type II TGF-b receptor in mouse skin results in an increase in carcinoma incidence and an acceleration of carcinoma development. Oncogene 17, 25–34. [107] Lu, S.L., Herrington, H., Reh, D., Weber, S., Bornstein, S., Wang, D., Li, A.G., Tang, C.F., Siddiqui, Y., Nord, J., Andersen, P., Corless, C.L. and Wang, X.J. (2006) Loss of transforming growth factor-b type II receptor promotes metastatic head-and-neck squamous cell carcinoma. Genes Dev. 20, 1331– 1342. [108] Guasch, G., Schober, M., Pasolli, H.A., Conn, E.B., Polak, L. and Fuchs, E. (2007) Loss of TGFb signaling destabilizes homeostasis and promotes squamous cell carcinomas in stratified epithelia. Cancer Cell 12, 313–327. [109] Bhowmick, N.A., Chytil, A., Plieth, D., Gorska, A.E., Dumont, N., Shappell, S., Washington, M.K., Neilson, E.G. and Moses, H.L. (2004) TGF-b signaling in fibroblasts modulates the oncogenic potential of adjacent epithelia. Science 303, 848–851. [110] Birchmeier, C., Birchmeier, W., Gherardi, E. and Vande Woude, G.F. (2003) Met, metastasis, motility and more. Nat. Rev. Mol. Cell Biol. 4, 915–925. [111] Cheng, N., Chytil, A., Shyr, Y., Joly, A. and Moses, H.L. (2007) Enhanced hepatocyte growth factor signaling by type II transforming growth factor-b receptor knockout fibroblasts promotes mammary tumorigenesis. Cancer Res. 67, 4869–4877. [112] Yang, L., Huang, J., Ren, X., Gorska, A.E., Chytil, A., Aakre, M., Carbone, D.P., Matrisian, L.M., Richmond, A., Lin, P.C. and Moses, H.L. (2008) Abrogation of TGF b signaling in mammary carcinomas recruits Gr-1+CD11b+ myeloid cells that promote metastasis. Cancer Cell 13, 23–35. [113] Bierie, B., Stover, D.G., Abel, T.W., Chytil, A., Gorska, A.E., Aakre, M., Forrester, E., Yang, L., Wagner, K.U. and Moses, H.L. (2008) Transforming growth factorb regulates mammary carcinoma cell survival and interaction with the adjacent microenvironment. Cancer Res. 68, 1809–1819. [114] Bierie, B., Chung, C.H., Parker, J.S., Stover, D.G., Cheng, N., Chytil, A., Aakre, M., Shyr, Y. and Moses, H.L. (2009) Abrogation of TGF-b signaling enhances chemokine production and correlates with prognosis in human breast cancer. J. Clin. Invest. 119, 1571–1582. [115] Oft, M., Peli, J., Rudaz, C., Schwarz, H., Beug, H. and Reichmann, E. (1996) TGFb1 and Ha-Ras collaborate in modulating the phenotypic plasticity and invasiveness of epithelial tumor cells. Genes Dev. 10, 2462–2477. [116] Oft, M., Heider, K.H. and Beug, H. (1998) TGFb signaling is necessary for carcinoma cell invasiveness and metastasis. Curr. Biol. 8, 1243–1252. [117] Janda, E., Lehmann, K., Killisch, I., Jechlinger, M., Herzig, M., Downward, J., Beug, H. and Grunert, S. (2002) Ras and TGFb cooperatively regulate epithelial cell plasticity and metastasis: dissection of Ras signaling pathways. J. Cell Biol. 156, 299–313. [118] Waerner, T., Alacakaptan, M., Tamir, I., Oberauer, R., Gal, A., Brabletz, T., Schreiber, M., Jechlinger, M. and Beug, H. (2006) ILEI: a cytokine essential for EMT, tumor formation, and late events in metastasis in epithelial cells. Cancer Cell 10, 227–239. [119] Bertran, E., Caja, L., Navarro, E., Sancho, P., Mainez, J., Murillo, M.M., Vinyals, A., Fabra, A. and Fabregat, I. (2009) Role of CXCR4/SDF-1 a in the migratory phenotype of hepatoma cells that have undergone epithelial-mesenchymal

1970

[120]

[121]

[122]

[123]

[124]

[125]

[126]

[127]

[128]

[129]

[130] [131]

C.-H. Heldin et al. / FEBS Letters 586 (2012) 1959–1970 transition in response to the transforming growth factor-b. Cell. Signal. 21, 1595–1606. Toh, B., Wang, X., Keeble, J., Sim, W.J., Khoo, K., Wong, W.C., Kato, M., PrevostBlondel, A., Thiery, J.P. and Abastado, J.P. (2011) Mesenchymal transition and dissemination of cancer cells is driven by myeloid-derived suppressor cells infiltrating the primary tumor. PLoS Biol. 9, e1001162. de Graauw, M., van Miltenburg, M.H., Schmidt, M.K., Pont, C., Lalai, R., Kartopawiro, J., Pardali, E., Le Devedec, S.E., Smit, V.T., van der Wal, A., Van’t Veer, L.J., Cleton-Jansen, A.M., ten Dijke, P. and van de Water, B. (2010) Annexin A1 regulates TGF-b signaling and promotes metastasis formation of basal-like breast cancer cells. Proc. Natl. Acad. Sci. USA 107, 6340–6345. Wendt, M.K., Taylor, M.A., Schiemann, B.J. and Schiemann, W.P. (2011) Down-regulation of epithelial cadherin is required to initiate metastatic outgrowth of breast cancer. Mol. Biol. Cell 22, 2423–2435. Giampieri, S., Manning, C., Hooper, S., Jones, L., Hill, C.S. and Sahai, E. (2009) Localized and reversible TGFb signalling switches breast cancer cells from cohesive to single cell motility. Nat. Cell Biol. 11, 1287–1296. Giampieri, S., Pinner, S. and Sahai, E. (2010) Intravital imaging illuminates transforming growth factor b signaling switches during metastasis. Cancer Res. 70, 3435–3439. Padua, D., Zhang, X.H., Wang, Q., Nadal, C., Gerald, W.L., Gomis, R.R. and Massagué, J. (2008) TGFb primes breast tumors for lung metastasis seeding through angiopoietin-like 4. Cell 133, 66–77. Serganova, I., Moroz, E., Vider, J., Gogiberidze, G., Moroz, M., Pillarsetty, N., Doubrovin, M., Minn, A., Thaler, H.T., Massagué, J., Gelovani, J. and Blasberg, R. (2009) Multimodality imaging of TGFb signaling in breast cancer metastases. FASEB J. 23, 2662–2672. Korpal, M., Yan, J., Lu, X., Xu, S., Lerit, D.A. and Kang, Y. (2009) Imaging transforming growth factor-b signaling dynamics and therapeutic response in breast cancer bone metastasis. Nat. Med. 15, 960–966. Kang, Y., Siegel, P.M., Shu, W., Drobnjak, M., Kakonen, S.M., Cordon-Cardo, C., Guise, T.A. and Massagué, J. (2003) A multigenic program mediating breast cancer metastasis to bone. Cancer Cell 3, 537–549. Kang, Y., He, W., Tulley, S., Gupta, G.P., Serganova, I., Chen, C.R., ManovaTodorova, K., Blasberg, R., Gerald, W.L. and Massagué, J. (2005) Breast cancer bone metastasis mediated by the Smad tumor suppressor pathway. Proc. Natl. Acad. Sci. USA 102, 13909–13914. Yingling, J.M., Blanchard, K.L. and Sawyer, J.S. (2004) Development of TGF-b signalling inhibitors for cancer therapy. Nat. Rev. Drug Discov. 3, 1011–1022. Hawinkels, L.J.A.C. and ten Dijke, P. (2011) Exploring anti-TGF-b therapies in cancer and fibrosis. Growth Factors 29, 140–152.

[132] Muraoka, R.S., Dumont, N., Ritter, C.A., Dugger, T.C., Brantley, D.M., Chen, J., Easterly, E., Roebuck, L.R., Ryan, S., Gotwals, P.J., Koteliansky, V. and Arteaga, C.L. (2002) Blockade of TGF-b inhibits mammary tumor cell viability, migration, and metastases. J. Clin. Invest. 109, 1551–1559. [133] Yang, Y.A., Dukhanina, O., Tang, B., Mamura, M., Letterio, J.J., MacGregor, J., Patel, S.C., Khozin, S., Liu, Z.Y., Green, J., Anver, M.R., Merlino, G. and Wakefield, L.M. (2002) Lifetime exposure to a soluble TGF-b antagonist protects mice against metastasis without adverse side effects. J. Clin. Invest. 109, 1607–1615. [134] Biswas, S., Guix, M., Rinehart, C., Dugger, T.C., Chytil, A., Moses, H.L., Freeman, M.L. and Arteaga, C.L. (2007) Inhibition of TGF-b with neutralizing antibodies prevents radiation-induced acceleration of metastatic cancer progression. J. Clin. Invest. 117, 1305–1313. [135] Nam, J.S., Terabe, M., Mamura, M., Kang, M.J., Chae, H., Stuelten, C., Kohn, E., Tang, B., Sabzevari, H., Anver, M.R., Lawrence, S., Danielpour, D., Lonning, S., Berzofsky, J.A. and Wakefield, L.M. (2008) An anti-transforming growth factor b antibody suppresses metastasis via cooperative effects on multiple cell compartments. Cancer Res. 68, 3835–3843. [136] Reka, A.K., Kurapati, H., Narala, V.R., Bommer, G., Chen, J., Standiford, T.J. and Keshamouni, V.G. (2010) Peroxisome proliferator-activated receptor-c activation inhibits tumor metastasis by antagonizing Smad3-mediated epithelial-mesenchymal transition. Mol. Cancer Ther. 9, 3221–3232. [137] Doi, S., Zou, Y., Togao, O., Pastor, J.V., John, G.B., Wang, L., Shiizaki, K., Gotschall, R., Schiavi, S., Yorioka, N., Takahashi, M., Boothman, D.A. and Kuroo, M. (2011) Klotho inhibits transforming growth factor-b1 (TGF-b1) signaling and suppresses renal fibrosis and cancer metastasis in mice. J. Biol. Chem. 286, 8655–8665. [138] Melisi, D., Ishiyama, S., Sclabas, G.M., Fleming, J.B., Xia, Q., Tortora, G., Abbruzzese, J.L. and Chiao, P.J. (2008) LY2109761, a novel transforming growth factor b receptor type I and type II dual inhibitor, as a therapeutic approach to suppressing pancreatic cancer metastasis. Mol. Cancer Ther. 7, 829–840. [139] Ganapathy, V., Ge, R., Grazioli, A., Xie, W., Banach-Petrosky, W., Kang, Y., Lonning, S., McPherson, J., Yingling, J.M., Biswas, S., Mundy, G.R. and Reiss, M. (2010) Targeting the transforming growth factor-b pathway inhibits human basal-like breast cancer metastasis. Mol. Cancer 9, 122. [140] Connolly, E.C., Saunier, E.F., Quigley, D., Luu, M.T., De Sapio, A., Hann, B., Yingling, J.M. and Akhurst, R.J. (2011) Outgrowth of drug-resistant carcinomas expressing markers of tumor aggression after long-term TbRI/II kinase inhibition with LY2109761. Cancer Res. 71, 2339–2349.