• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptNIH Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Cancer Res. Author manuscript; available in PMC Jan 1, 2012.
Published in final edited form as:
PMCID: PMC3025803

Snail2 is an essential mediator of Twist1-induced epithelial-mesenchymal transition and metastasis


To metastasize, carcinoma cells must attenuate cell-cell adhesion to disseminate into distant organs. A group of transcription factors, including Twist1, Snail1, Snail2, ZEB1, and ZEB2, have been shown to induce Epithelial-Mesenchymal Transition (EMT), thus promoting tumor dissemination. However, it is unknown whether these transcription factors function independently or coordinately to activate the EMT program. Here we report that direct induction of Snail2 is essential for Twist1 to induce EMT. Snail2 knockdown completely blocks the ability of Twist1 to suppress E-cadherin transcription. Twist1 binds to an evolutionarily conserved E-box on the proximate Snail2 promoter to induce its transcription. Snail2 induction is essential for Twist1-induced cell invasion and distant metastasis in mice. In human breast tumors, the expression of Twist1 and Snail2 is highly correlated. Together, our results show that Twist1 needs to induce Snail2 to suppress the epithelial branch of the EMT program and that Twist1 and Snail2 act together to promote EMT and tumor metastasis.

Keywords: tumor metastasis, E-cadherin, Epithelial-Mesenchymal Transition, Snail2, Twist1


For carcinoma cells to break away from neighboring cells and invade, they must lose cell-cell adhesion and gain motility (1). A highly conserved developmental program named Epithelial-Mesenchymal Transition (EMT) has been implicated in promoting dissemination of single carcinoma cells from primary epithelial tumors (2). The EMT program is activated during many developmental processes–such as mesoderm formation and neural crest development (3). During EMT, cells lose their epithelial characteristics, including cell adhesion and polarity, and acquire a mesenchymal morphology and the ability to migrate. Biochemically, cells switch off the expression of epithelial markers, such as adherens junction protein E-cadherin and turn on mesenchymal markers, including vimentin and fibronectin (4).

Functional loss of E-cadherin in an epithelial cell has been considered a hallmark of EMT. Forced expression of E-cadherin in certain invasive carcinoma cells can inhibit their ability to invade and metastasize; conversely, blocking E-cadherin function in non-invasive tumor cells activates their invasiveness and metastatic abilities (5, 6). A partial loss of E-cadherin is also associated with carcinoma progression and poor prognosis in various tumor types (7).

Transcriptional modulation of the E-cadherin gene promoter is a key event to suppress E-cadherin expression during EMT. The human E-cadherin promoter contains E-box elements that are responsible for its transcriptional repression (8, 9). Several Zn-finger transcription factors, including Snail1 (10, 11), Snail2 (12), ZEB1 (13), and ZEB2 (14), are capable of directly binding to the E-boxes of the E-cadherin promoter to repress its transcription. In addition to these Zn-finger transcription factors, transcription factors belonging to other families have also been shown to be able to regulate EMT in culture and during development. In a search for genes involved in mouse mammary tumor metastasis, the bHLH transcription factor Twist1 is found to be capable of inducing EMT in human mammary epithelial cells. Our previous study also found that the Twist1 transcription factor was essential for the ability of tumor cells to metastasize from the mammary gland to the lung in a mouse breast tumor model (15).

During EMT in metastasis and in embryogenesis, many EMT-inducing transcription factors are often activated simultaneously, such as expression of Twist1, Snail1, Snail2 and ZEB2 in neural crest cells (16, 17). To understand how these transcription factors coordinate the EMT program, we have used an inducible Twist1 system to address whether and how Twist1 activates other EMT-inducing transcription factors to suppress E-cadherin and promote EMT and tumor metastasis.

Materials and methods

Cell lines

HMLE cells were obtained from Dr. Robert Weinberg and cultured as described (18). SUM1315 and MCF7 cells were obtained from ATCC and cultured as described (19).


Primary antibodies used include Twist1 (20), E-cadherin, β-catenin, γ-catenin, fibronectin, vimentin, N-cadherin (BD Biosciences), β-actin (Abcam), as previously described (15); H-Ras (18), Snail2 (21) (Santa Cruz), and α-tubulin (22) (Abcam).

Quantitative PCR

Total RNAs were extracted using RNeasy Mini Kit (Qiagen) and reverse transcribed using cDNA Reverse Transcription Kit (Applied Biosystems). Resulting cDNAs were analyzed in triplicates using SYBR-Green PCR mix (Applied Biosystems). Relative mRNA concentrations were determined by 2−(Ct-Cc) where Ct and Cc are the mean threshold cycle differences after normalizing to GAPDH values. Primers used for PCR are listed in Supplementary Materials.


Cells were grown on coverslip slides for 3 days, fixed with 4% paraformaldehyde, permeabilized with 0.1%Triton X-100 for 10 min, and blocked with 5% goat serum. Samples were incubated with primary and Alexa secondary antibodies (Invitrogen). After washing, wells were covered with SlowFade with DAPI (Invitrogen). Antibodies were used as described (15).

Chip sequencing

An Illumina sequencing library was prepared from Twist1 ChIP DNA obtained from induced HMLE-Twist1-ER cells using the ChIP-seq Sample Preparation Kit (Illumina). The DNA library was sequenced on the Illumina Genome Analyzer. The resulting sequence reads were mapped to the human genome using the Illumina software suite. The data are displayed on the UCSC genome browser 2006 assembly.

Chromatin Immunoprecipitation

Cells were crosslinked with 1% Paraformaldehyde (PFA), lysed and sonicated. Nuclear lysates were incubated with Protein G Dynabeads (Invitrogen) conjugated with an anti-Twist1 or anti-estrogen receptor antibody overnight. DNA was reverse crosslinked and purified. Primers used for PCR were described in Supplementary Materials.

Luciferase Reporter assay

MCF7 cells were transfected with Snail2prom-Luc2 reporter plasmid, pGL4[Rluc] plasmid, and additional plasmids as described. 24 hours later, the cell lysates were assayed accordingly (Promega). The firefly luciferase activity was normalized to that of Renilla luciferase.

Promoter sequence alignment

Analysis of the Snail2 promoter E-box sequences across species was performed using ConTra online software (23).

Invasion and Migration Assays

40,000 cells were cultured on 8mm Transwell (Costar) for 72 hours, fixed with 4% paraformaldehyde, washed and stained with 0.1% crystal violet. The top membrane was cleaned, washed, and dried. Crystal violet was released with 10% acetic acid and the absorbency was measured at 520 nm. All assays were performed in triplicates.

Subcutaneous tumor implantation and metastasis assay

1 million of HMLER-Twist1 cells expressing shControl and 4 million of HMLER-Twist1 cells expressing shSnail2 were injected subcutaneously into Nude mice. Tumor size was measured weekly until reaching 2cm in diameter before mice were sacrificed. Lungs were imaged under a fluorescence dissection microscope. The total number of GFP positive colonies was quantified using ImageJ. Graphed tumor weight represents total primary tumor burden per mouse.

Bioinformatics and Statistical Analysis

Three published microarray datasets were downloaded from NCBI’s Gene Expression Omnibus (GEO). Obtained GEO identifiers (Study name, sample size) were GSE1456 (STOCKHOLM, n=159), GSE2034 (EMC, n=286), and GSE3494 (UPPSALA, n=236). All samples were from breast cancer patients. Quantile-normalization was performed on an integrated dataset of .CEL files (24). The Median-Polish algorithm was used to convert multiple probe-level measurement values into the gene-level aggregate value (25). The Spearman correlation coefficient was calculated between Twist1 and individual genes and ranked in individual datasets. Measurement values of the two probe sets were plotted in Supplementary Figure 4 with a linear regression model fit.


Twist1 indirectly suppresses E-cadherin transcription to promote EMT

To understand how Twist1 suppresses E-cadherin expression during EMT, we have generated a fusion protein between Twist1 and the modified hormone-binding domain of estrogen receptor (Twist1-ER), which can be activated specifically upon binding to 4-hydroxytamoxifen (26). We expressed Twist1-ER in the immortalized human mammary epithelial (HMLE) cells via retroviral infection. HMLE cells expressing Twist1-ER underwent morphological changes of EMT upon 4-hydroxy-tamoxifen treatment in 10–12 days (Figure 1A). We harvested these cells at day 18 to characterize the EMT markers and found that activation of Twist1 resulted in loss of cell adhesion and gain of mesenchymal differentiation in HMLE cells (Figure 1B), identical to the EMT program induced by Twist1 (15).

Figure 1
Twist1 indirectly suppresses E-cadherin transcription to promote EMT

We next examined the expression of E-cadherin mRNA and protein following Twist1 activation. Interestingly, both E-cadherin mRNA and protein levels remained unchanged during the first 7 days after Twist1 activation and only started decreasing afterwards. By the end of the 12 day time course, both E-cadherin mRNA and protein decreased drastically (Figure 1C and 1D), which correlated with loss of cell adhesion in cell morphology (Figure 1A). Given that 4-hydroxytamoxifen activates the Twist1-ER fusion protein immediately (26), the 6–8 days required to reduce E-cadherin mRNA and protein levels suggest that Twist1 does not directly suppress E-cadherin transcription during EMT.

Induction of the transcription factor Snail2 is required for Twist1-induced EMT

Since Twist1 may not directly suppress E-cadherin transcription, we hypothesized that Twist1 might activate other Zn-finger transcription factors such as Snail1, Snail2, ZEB1 and ZEB2 to suppress E-cadherin transcription. During Drosophila mesoderm formation, Twist1 directly induces the expression of Snail1 to promote EMT (27). To test whether Twist1 induces the Snail family transcription factors in human cells, we examined Snail1 and Snail2 expression following Twist1 activation. Interestingly, Snail2 was induced within four hours upon Twist1 activation (Figure 2A), while Snail1 was only significantly induced at Day 12 (Supplementary Figure 1). Induction of Snail2 protein was also observed in HMLE-Twist1-ER cells two days upon Twist1 activation and in HMLE cells expressing Twist1 (Figure 2B). Since we are interested in possible direct target genes that Twist1 regulates to suppress E-cadherin, we hypothesized that induction of Snail2 by Twist1 could play a key role in the transcriptional repression of E-cadherin.

Figure 2
Induction of the transcription factor Snail2 is required for Twist1-induced EMT

To test whether induction of Snail2 is required for the ability of Twist1 to induce EMT, we designed two shRNA lentiviral constructs specifically against human Snail2 and expressed them in HMLE cells. Both shSnail2-1 and shSnail2-2 could efficiently suppress Snail2 mRNA and protein below the basal level in HMLE cells (Figure 2A–2C). We expressed both shRNAs against Snail2 in HMLE cells expressing the inducible Twist1-ER construct and treated the cells with 4-hydroxy-tamoxifen. Interestingly, in the absence of Snail2, Twist1 was not able to induce the morphological changes associated with EMT as observed in HMLE cells expressing a control shRNA construct. Instead, these cells with shSnail2 constructs maintained an epithelial morphology with intact cell-cell adhesions and a cobblestone-like morphology even after 18 days (Figure 2D). Also knockdown of Snail2 did not affect Twist1 expression in these cells (Figure 2B). Together, these data indicate that induction of Snail2 is required for the ability of Twist1 to promote EMT in HMLE cells.

Induction of Snail2 is specifically responsible for suppression of E-cadherin in response to Twist1 activation

To determine how suppression of Snail2 affects expression of various EMT markers, we closely followed the entire time course of Twist1 activation in HMLE-Twist1-ER cells expressing shSnail2 or a control shRNA. First, we examined the expression of E-cadherin following the induction time course, since Snail2 is a direct repressor of E-cadherin (12). Indeed, suppression of Snail2 in HMLE-Twist1-ER cells completely inhibited the progressive reduction of E-cadherin mRNA and protein upon 4-hydroxytamoxifen treatment (Figure 3A and 3B). In HMLE-Twist1-ER cells expressing a control shRNA, E-cadherin expression started decreasing at Day 7 after Twist1 activation; at Day 18, E-cadherin protein was completely lost and β-catenin disappeared from cell membranes, indicating the loss of adherens junctions upon Twist1 activation (Figure 3C). In contrast, in HMLE-Twist1-ER cells expressing shSnail2, both E-cadherin and β-catenin remained at cell-cell junctions even after 18 days of Twist1 activation (Figure 3C). These results indicate that Snail2 is critical for the ability of Twist1 to suppress E-cadherin and induce EMT.

Figure 3
Induction of Snail2 is required for suppression of E-cadherin during Twist1-induced EMT

To understand whether suppression of E-cadherin by Snail2 is the only effect of Snail2 on the entire EMT program, we examined the mRNA and protein expression of EMT mesenchymal markers in response to Twist1 activation. Interestingly, we found that unlike E-cadherin, the initial induction of other EMT markers, including fibronectin, vimentin, and N-cadherin occurred normally in the absence of Snail2 in HMLE-Twist1-ER cells upon 4-hydroxytamoxifen treatment (Figure 4A). However, at Day 7 when E-cadherin was not suppressed in HMLE-Twist1-ER cells expressing shSnail2, fibronectin, vimentin, and N-cadherin proteins failed to increase further to reach the high levels of induction observed in HMLE-Twist1-ER cells expressing a control shRNA (Figure 4A and 4B). We also performed immunofluorescence analysis to examine the expression and localization of these EMT markers and observed similar effects (Supplementary Figure 2). In addition, we attempted to overexpress Snail2 in HMLE cells and found that Snail2 alone was not sufficient to induce EMT in HMLE cells (data not shown). These results suggest that, although Snail2 is not directly responsible for the induction of fibronectin, vimentin, and N-cadherin during EMT, loss of Snail2 and E-cadherin suppression completely blocks EMT morphogenesis and attenuates the expression of mesenchymal markers, including fibronectin, vimentin, and N-cadherin.

Figure 4
Suppression of Snail2 attenuates expression of mesenchymal markers during Twist1-induced EMT

Snail2 is a conserved direct transcriptional target of Twist1 in amniotes

Since Snail2 is induced immediately upon Twist1 activation and plays a critical role in Twist1-induced EMT, we tested whether Twist1 directly binds to the Snail2 gene promoter to activate its transcription. Using Chromatin Immunoprecipitation (ChIP) coupled with high-throughput sequencing, we identified a distinct Twist1 binding E-box domain 306bp from the transcription start site of the human Snail2 promoter (Figure 5A). In contrast, there is no significant enrichment for the 10kb promoter regions of Snail1or E-cadherin (Figure 5A). To demonstrate that Twist1 specifically binds to the Snail2 promoter, we treated HMLE-Twist1-ER cells with 4-hydroxytamoxifen for four days and performed ChIP analysis to pull-down Twist1-binding DNA fragments. Both Twist1 and ER antibodies were able to specifically enrich the E-box region of the human Snail2 promoter, while there was no specific enrichment for the control GAPDH gene (Figure 5B). To test the specificity of the Twist1-binding E-box on the Snail2 promoter, we tested the ability of Twist1 to activate the isolated 606bp human Snail2 promoter by luciferase reporter assay. We found that Twist1 was able to activate the Snail2 promoter-driven luciferase activity and mutation of the E-box sequence on the Snail2 promoter reduced Twist1-reduced activation (Figure 5C). These data suggest that Twist1 specifically and directly binds to the E-box on the human Snail2 promoter to activate its transcription.

Figure 5
Snail2 is an evolutionally conserved direct transcriptional target of Twist1 in amniotes

To test whether the direct regulation of Snail2 by Twist1 is conserved throughout evolution, we searched for the consensus Twist1-binding E-box and its surrounding sequence on the human Snail2 promoter across species. Interestingly, we found that this consensus Twist1-binding sequence (ccaCAtcTGgaagcc) on the Snail2 promoter is highly conserved among amniote genomes examined, including chimpanzee, monkey, mice, rabbit, armadillo, elephant, hedgehog, and chick (Figure 5D). However, this sequence could not be found in Xenopus and Zebrafish, indicating that the direct induction of Snail2 by Twist1 is highly conserved in amniotes.

Induction of Snail2 is required for the ability of Twist1 to promote invasion and metastasis

Given that induction of Snail2 is required for activating the EMT program by Twist1, we next tested whether Snail2 plays an essential role in Twist1-mediated tumor cell invasion and metastasis. As in HMLE-Twist1-ER cells, knockdown of Snail2 also prevented constitutive Twist1-induced EMT in HMLE cells (Figure 6A). We tested whether Snail2 is required for Twist1-induced cell migration and invasion. As shown in Figure 6B, suppression of Snail2 resulted in reduced cell migration and invasion. This data indicate that cell migration and invasion induced by Twist1 require the activation of the entire EMT program. Without Snail2, Twist1 is not sufficient to promote cell migration and invasion.

Figure 6
Induction of Snail2 is required for the ability of Twist1 to promote invasion and metastasis

We next tested whether Snail2 is required for the ability of Twist1 to promote tumor metastasis in vivo. To assess the role of Snail2 in tumor metastasis, we transformed GFP-labeled HMLE-Twist1 cells expressing shSnail2 or a control shRNA with oncogenic Ras (HMLER cells) (Figure 6A). When we implanted these cells subcutaneously in nude mice, HMLER-Twist1 cells expressing a control shRNA generated hundreds of tiny GFP+ micrometastatic lesions in the lung (Figure 6C). Interestingly, when Snail2 was knocked down, HMLER-Twist1 cells failed to form GFP+ metastatic lesions in the lung (Figure 6C), even though the primary tumors reached similar weight as shControl tumors (Figure 6D). It is worth noting that knockdown Snail2 reduced the initial primary tumor seeding rates, probably due to the difficulty for epithelial cells to establish cell-cell adhesion after being injected subcutaneously in suspension. Therefore we allowed the Snail2 knockdown tumors to grow four more weeks in mice to reach the same primary tumor size as the control tumors (Supplementary Figure 3). Even with longer primary tumor growth time and similar final primary tumor burdens, examination of individual lung lobes revealed that suppression of Snail2 caused a 500-fold reduction in the ability of HMLER-Twist1 cells to form micrometastasis in the lung (Figure 6D). Immunohistochemistry analysis of the resulting tumors showed that E-cadherin expression was completely absent from cell membranes in HMLE-Twist1 tumors, while suppression of Snail2 resulted in epithelial tumors with strong E-cadherin expression at cell-cell junctions (Figure 6C). Together, these results demonstrate that activation of Twist1 is sufficient, at least in some cell types, to promote tumor cells to disseminate from the primary site to distant organs. Furthermore, the data strongly indicate that Snail2 functions downstream of Twist1 to play a critical role in regulating tumor invasion and metastasis.

Twist1 and Snail2 are frequently co-expressed in human breast tumors

To further demonstrate that a direct link between Twist1 and Snail2 exists in human tumor cells, we expressed shRNAs against Twist1 in the human breast tumor cell line SUM1315 that expresses high levels of endogenous Twist1 and Snail2. Indeed, when we examined the expression of endogenous Snail2 in these cells, we observed a dramatic reduction of Snail2 expression upon suppression of Twist1 (Figure 7A). This result indicates that expression of endogenous Snail2 is dependent on Twist1 in human breast tumor cells.

Figure 7
Twist1 and Snail2 are frequently co-expressed in human breast tumors

To more directly probe the association between Twist1 and Snail2 in human breast tumor samples, we analyzed three published human breast tumor gene expression datasets summing up to 681 primary breast cancers (2830). In each data set, we ranked all 22282 genes based on how tight their expression levels are correlated with Twist1 in individual samples. As shown in Figure 7B, in all three datasets, Snail2 was consistently ranked as one of the most correlated genes with Twist1. Furthermore, expression of Twist1 and Snail2 were positively correlated (Supplementary Figure 4). The association between Twist1 and Snail2 in human breast tumor samples provides further support for a functional link between Twist1 and Snail2 in breast tumor progression.


Our study has identified Snail2 as the direct transcription target of Twist1 in repressing E-cadherin during the EMT program. We show that induction of Snail2 is required for the ability of Twist1 to promote EMT in culture and distant metastasis in mice. Furthermore, we identified a conserved Twist1-binding site on the proximate Snail2 promoter in amniotes. These results show that Twist1 induces Snail2 to coordinate the EMT program to promote tumor metastasis (Figure 7C).

Our study revealed an evolutionally conserved link between Twist1 and its direct transcription target Snail2 in suppressing E-cadherin transcription, thus promoting the loss of cell adhesion during EMT. In Drosophila, Twist1 directly induces Snail1 to promote EMT during mesoderm formation (27). Our study shows that in humans, Snail2, a family member, replaces Snail1 to perform this key function downstream of Twist1 during EMT. Elevated expression of both Twist1 and Snail2 have been observed in lymph node-positive breast tumors and associated with poor outcome (31). Twist1 and Snail2 are also functionally linked during early embryogenesis. In early Xenopus embryo, loss of Twist1 leads to a decrease in the Snail2 mRNA level, and both of them negatively regulate the expression of Cerberus, an inhibitor of Nodal, BMP, and Wnt signaling (32). During Xenopus embryonic vasculogenesis, knockdown of Twist1 or Snail2 provoked similar embryonic hemorrhage, and either Twist1 or Snail2 could rescue the hemorrhage defects due to Myc knockdown (33). Together, these date show that the direct link between Twist1 and Snail2 is highly conserved and functionally important during embryogenesis and tumor metastasis.

Interestingly, induction of Snail2 is not only required for the ability of Twist1 to suppress E-cadherin, but the entire EMT morphogenesis is blocked in the absence of Snail2. Further analyses of additional EMT markers following the time course of Twist1 activation indicate that genes such as N-cadherin are able to initially increase at similar rates in the absence of Snail2. The initial immediate induction of N-cadherin is possibly due to the direct action of Twist1, together with its dimerization partner E12/E47. Interestingly, N-cadherin failed to further increase to the high induction level after 7 days of Twist1 activation, the exact time when E-cadherin is supposed to be suppressed upon Twist1 activation. This result indicates that although Snail2 is not directly responsible for the induction of N-cadherin, induction of Snail2 and suppression of E-cadherin are essential to send out a permissive signal to allow further expression of mesenchymal markers and the progression of the EMT program.

The permissive signal that allows progression of the EMT program remains unknown. The disappearance of E-cadherin from adherens junctions results in the release of its partner, β-catenin, into the cytosol, which has the potential to enter the nucleus to activate LEF/TCF-mediated transcription. Although the contribution of the adherens junction-associated β-catenin pool to transcription is still a matter of debate, several studies suggest that β-catenin-mediated transcription can further induce the expression of Snail2 (34) and Twist1 (35), thereby contributing to the EMT program. Furthermore, fibronectin and vimentin have been shown to be β-catenin target genes (3638), which explains why suppression of Snail2 expression blocks their further induction in response to Twist1 activation. Similarly, Zeb2 has also been shown to not only suppress E-cadherin, but also partially regulate vimentin expression (39). Together, these studies suggest that suppression of E-cadherin not only breaks the physical barrier to allow cell scattering, but also could actively contribute to activation of mesenchymal traits during EMT.

We found that Snail2 was induced four hours after Twist1 activation; however, reduction of E-cadherin mRNA only became significant seven days after Twist1 activation. This observation raises the question of whether Snail2 alone is sufficient to suppress E-cadherin transcription during Twist1-induced EMT. Several observations could help to explain this phenomenon. First, we found that two additional E-cadherin suppressing transcription factors, ZEB1 and ZEB2, were also induced upon Twist1 activation; and suppression of Snail2 drastically inhibited induction of ZEB1 and ZEB2 upon Twist1 activation. Therefore, induction of ZEB1 and ZEB2 could be required to collaborate with Snail2 to achieve drastic suppression of E-cadherin transcription. A double-negative feedback loop controlling ZEB1/ZEB2 and microRNA-200 family expression could further promote the expression of ZEB1 and ZEB2 during EMT (4043). Together, these factors strengthen the ability of cells to suppress E-cadherin expression during EMT. Second, recent studies found that Ajuba LIM proteins, also components of adherens junctions, can travel to the nucleus to serve as corepressors with the Snail family transcription factors to repress E-cadherin (44, 45). Therefore, suppression of E-cadherin during EMT progresses in a positive feedback loop. At the beginning of EMT, a slight decrease of E-cadherin causes the weakening of cell adhesion and the release of Ajuba LIM proteins; binding to Snail2 or Snail1, Ajuba LIM proteins can then further suppress E-cadherin transcription. These transcription repressors, including Snail1/Snail2, Ajuba LIM proteins, ZEB1/ZEB2, and the microRNA-200 family, function as a powerful E-cadherin suppression machine to drastically suppress E-cadherin transcription and promote EMT progression.

Supplementary Material


We are grateful to members of the Yang lab, especially Etienne Danis and Andrew T. Chang, for their help and suggestions. We thank Dr. Xiang-Dong Fu and his lab for their help with ChIP-Sequencing. We thank Dr. Cornelis Murre for providing the E47 expression construct.

Grant Support

This work was supported by grants from American Cancer Society (RSG-09-282-01-CSM), Kimmel Scholar Award, California Breast Cancer Research Program (12IB-0065), and the Mary Kay Ash Charitable Foundation (J.Y.), by the UCSD Graduate Training Program in Cellular and Molecular Pharmacology from the National Institute of General Medical Sciences T32 GM007752 and a DOD Breast Cancer Predoctoral fellowship (E.C.), and by Susan G. Komen Foundation grant FAS0703850 (J.K. and L.O.).


Disclosure of Potential Conflict of interest

No potential conflict of interest was disclosed.


1. Fidler IJ. The pathogenesis of cancer metastasis: the 'seed and soil' hypothesis revisited. Nat Rev Cancer. 2003;3:453–8. [PubMed]
2. Thiery JP. Epithelial-mesenchymal transitions in tumour progression. Nat Rev Cancer. 2002;2:442–54. [PubMed]
3. Yang J, Weinberg RA. Epithelial-mesenchymal transition: at the crossroads of development and tumor metastasis. Dev Cell. 2008;14:818–29. [PubMed]
4. Hay ED. An overview of epithelio-mesenchymal transformation. Acta Anat (Basel) 1995;154:8–20. [PubMed]
5. Frixen UH, Behrens J, Sachs M, et al. E-cadherin-mediated cell-cell adhesion prevents invasiveness of human carcinoma cells. J Cell Biol. 1991;113:173–85. [PMC free article] [PubMed]
6. Perl AK, Wilgenbus P, Dahl U, Semb H, Christofori G. A causal role for E-cadherin in the transition from adenoma to carcinoma. Nature. 1998;392:190–3. [PubMed]
7. Vincent-Salomon A, Thiery JP. Host microenvironment in breast cancer development: epithelial-mesenchymal transition in breast cancer development. Breast Cancer Res. 2003;5:101–6. [PMC free article] [PubMed]
8. Giroldi LA, Bringuier PP, de Weijert M, Jansen C, van Bokhoven A, Schalken JA. Role of E boxes in the repression of E-cadherin expression. Biochem Biophys Res Commun. 1997;241:453–8. [PubMed]
9. Hennig G, Behrens J, Truss M, Frisch S, Reichmann E, Birchmeier W. Progression of carcinoma cells is associated with alterations in chromatin structure and factor binding at the E-cadherin promoter in vivo. Oncogene. 1995;11:475–84. [PubMed]
10. Batlle E, Sancho E, Franci C, et al. The transcription factor snail is a repressor of E- cadherin gene expression in epithelial tumour cells. Nat Cell Biol. 2000;2:84–9. [PubMed]
11. Cano A, Perez-Moreno MA, Rodrigo I, et al. The transcription factor snail controls epithelial-mesenchymal transitions by repressing E-cadherin expression. Nat Cell Biol. 2000;2:76–83. [PubMed]
12. Hajra KM, Chen DY, Fearon ER. The SLUG zinc-finger protein represses E-cadherin in breast cancer. Cancer Res. 2002;62:1613–8. [PubMed]
13. Eger A, Aigner K, Sonderegger S, et al. DeltaEF1 is a transcriptional repressor of E-cadherin and regulates epithelial plasticity in breast cancer cells. Oncogene. 2005;24:2375–85. [PubMed]
14. Comijn J, Berx G, Vermassen P, et al. The two-handed E box binding zinc finger protein SIP1 downregulates E-cadherin and induces invasion. Mol Cell. 2001;7:1267–78. [PubMed]
15. Yang J, Mani SA, Donaher JL, et al. Twist, a master regulator of morphogenesis, plays an essential role in tumor metastasis. Cell. 2004;117:927–39. [PubMed]
16. Nieto MA. The snail superfamily of zinc-finger transcription factors. Nat Rev Mol Cell Biol. 2002;3:155–66. [PubMed]
17. Van de Putte T, Maruhashi M, Francis A, et al. Mice lacking ZFHX1B, the gene that codes for Smad-interacting protein-1, reveal a role for multiple neural crest cell defects in the etiology of Hirschsprung disease-mental retardation syndrome. Am J Hum Genet. 2003;72:465–70. [PMC free article] [PubMed]
18. Elenbaas B, Spirio L, Koerner F, et al. Human breast cancer cells generated by oncogenic transformation of primary mammary epithelial cells. Genes Dev. 2001;15:50–65. [PMC free article] [PubMed]
19. Neve RM, Chin K, Fridlyand J, et al. A collection of breast cancer cell lines for the study of functionally distinct cancer subtypes. Cancer Cell. 2006;10:515–27. [PMC free article] [PubMed]
20. Gitelman I. Twist protein in mouse embryogenesis. Dev Biol. 1997;189:205–14. [PubMed]
21. Mittal MK, Myers JN, Misra S, Bailey CK, Chaudhuri G. In vivo binding to and functional repression of the VDR gene promoter by SLUG in human breast cells. Biochem Biophys Res Commun. 2008;372:30–4. [PMC free article] [PubMed]
22. Kilmartin JV, Wright B, Milstein C. Rat monoclonal antitubulin antibodies derived by using a new nonsecreting rat cell line. J Cell Biol. 1982;93:576–82. [PMC free article] [PubMed]
23. Hooghe B, Hulpiau P, van Roy F, De Bleser P. ConTra: a promoter alignment analysis tool for identification of transcription factor binding sites across species. Nucleic Acids Res. 2008;36:W128–32. [PMC free article] [PubMed]
24. Bolstad BM, Irizarry RA, Astrand M, Speed TP. A comparison of normalization methods for high density oligonucleotide array data based on variance and bias. Bioinformatics. 2003;19:185–93. [PubMed]
25. Irizarry RA, Bolstad BM, Collin F, Cope LM, Hobbs B, Speed TP. Summaries of Affymetrix GeneChip probe level data. Nucleic Acids Res. 2003;31:e15. [PMC free article] [PubMed]
26. Littlewood TD, Hancock DC, Danielian PS, Parker MG, Evan GI. A modified oestrogen receptor ligand-binding domain as an improved switch for the regulation of heterologous proteins. Nucleic Acids Res. 1995;23:1686–90. [PMC free article] [PubMed]
27. Leptin M. twist and snail as positive and negative regulators during Drosophila mesoderm development. Genes Dev. 1991;5:1568–76. [PubMed]
28. Miller LD, Smeds J, George J, et al. An expression signature for p53 status in human breast cancer predicts mutation status, transcriptional effects, and patient survival. Proc Natl Acad Sci U S A. 2005;102:13550–5. [PMC free article] [PubMed]
29. Pawitan Y, Bjohle J, Amler L, et al. Gene expression profiling spares early breast cancer patients from adjuvant therapy: derived and validated in two population-based cohorts. Breast Cancer Res. 2005;7:R953–64. [PMC free article] [PubMed]
30. Sotiriou C, Wirapati P, Loi S, et al. Gene expression profiling in breast cancer: understanding the molecular basis of histologic grade to improve prognosis. J Natl Cancer Inst. 2006;98:262–72. [PubMed]
31. Martin TA, Goyal A, Watkins G, Jiang WG. Expression of the transcription factors snail, slug, and twist and their clinical significance in human breast cancer. Ann Surg Oncol. 2005;12:488–96. [PubMed]
32. Zhang C, Klymkowsky MW. Unexpected functional redundancy between Twist and Slug (Snail2) and their feedback regulation of NF-kappaB via Nodal and Cerberus. Dev Biol. 2009;331:340–9. [PMC free article] [PubMed]
33. Rodrigues CO, Nerlick ST, White EL, Cleveland JL, King ML. A Myc-Slug (Snail2)/Twist regulatory circuit directs vascular development. Development. 2008;135:1903–11. [PMC free article] [PubMed]
34. Conacci-Sorrell M, Simcha I, Ben-Yedidia T, Blechman J, Savagner P, Ben-Ze'ev A. Autoregulation of E-cadherin expression by cadherin-cadherin interactions: the roles of beta-catenin signaling, Slug, and MAPK. J Cell Biol. 2003;163:847–57. [PMC free article] [PubMed]
35. Onder TT, Gupta BP, Mani SA, Yang J, Lander ES, Weinberg RA. Loss of E-cadherin promotes metastasis via multiple downstream transcriptional pathways. Cancer Research. 2008;68:3645–54. [PubMed]
36. Gradl D, Kuhl M, Wedlich D. The Wnt/Wg signal transducer beta-catenin controls fibronectin expression. Mol Cell Biol. 1999;19:5576–87. [PMC free article] [PubMed]
37. De Langhe SP, Sala FG, Del Moral PM, et al. Dickkopf-1 (DKK1) reveals that fibronectin is a major target of Wnt signaling in branching morphogenesis of the mouse embryonic lung. Dev Biol. 2005;277:316–31. [PubMed]
38. Gilles C, Polette M, Mestdagt M, et al. Transactivation of vimentin by beta-catenin in human breast cancer cells. Cancer Res. 2003;63:2658–64. [PubMed]
39. Bindels S, Mestdagt M, Vandewalle C, et al. Regulation of vimentin by SIP1 in human epithelial breast tumor cells. Oncogene. 2006;25:4975–85. [PubMed]
40. Gregory PA, Bert AG, Paterson EL, et al. The miR-200 family and miR-205 regulate epithelial to mesenchymal transition by targeting ZEB1 and SIP1. Nat Cell Biol. 2008;10:593–601. [PubMed]
41. Park SM, Gaur AB, Lengyel E, Peter ME. The miR-200 family determines the epithelial phenotype of cancer cells by targeting the E-cadherin repressors ZEB1 and ZEB2. Genes Dev. 2008;22:894–907. [PMC free article] [PubMed]
42. Bracken CP, Gregory PA, Kolesnikoff N, et al. A double-negative feedback loop between ZEB1-SIP1 and the microRNA-200 family regulates epithelial-mesenchymal transition. Cancer Res. 2008;68:7846–54. [PubMed]
43. Korpal M, Lee ES, Hu G, Kang Y. The miR-200 family inhibits epithelial-mesenchymal transition and cancer cell migration by direct targeting of E-cadherin transcriptional repressors ZEB1 and ZEB2. J Biol Chem. 2008;283:14910–4. [PMC free article] [PubMed]
44. Ayyanathan K, Peng H, Hou Z, et al. The Ajuba LIM domain protein is a corepressor for SNAG domain mediated repression and participates in nucleocytoplasmic Shuttling. Cancer Res. 2007;67:9097–106. [PubMed]
45. Langer EM, Feng Y, Zhaoyuan H, Rauscher FJ, 3rd, Kroll KL, Longmore GD. Ajuba LIM proteins are snail/slug corepressors required for neural crest development in Xenopus. Dev Cell. 2008;14:424–36. [PMC free article] [PubMed]
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...