Telomerase reverse transcriptase promotes epithelial&ndash ... - Nature

2 downloads 0 Views 4MB Size Report
Oct 8, 2012 - Telomerase activation through induction of telomerase reverse transcriptase (hTERT) contributes to malignant transformation by stabilizing ...
Oncogene (2013) 32, 4203–4213 & 2013 Macmillan Publishers Limited All rights reserved 0950-9232/13 www.nature.com/onc

ORIGINAL ARTICLE

Telomerase reverse transcriptase promotes epithelial–mesenchymal transition and stem cell-like traits in cancer cells Z Liu1,2, Q Li1,2, K Li3, L Chen1, W Li1, M Hou4, T Liu2, J Yang2, C Lindvall5, M Bjo¨rkholm2, J Jia1 and D Xu2,6 Telomerase activation through induction of telomerase reverse transcriptase (hTERT) contributes to malignant transformation by stabilizing telomeres. Clinical studies demonstrate that higher hTERT expression is associated with cancer progression and poor outcomes, but the underlying mechanism is unclear. Because epithelial–mesenchymal transition (EMT) and cancer stem cells (CSCs) are key factors in cancer metastasis and relapse, and hTERT has been shown to exhibit multiple biological activities independently of its telomere-lengthening function, we address a potential role of hTERT in EMT and CSCs using gastric cancer (GC) as a model. hTERT overexpression promotes, whereas its inhibition suppresses, EMT and stemness of GC cells, respectively. Transforming growth factor (TGF)-b1 and b-catenin-mediated EMT was abolished by small interfering RNA depletion of hTERT expression. hTERT interacts with b-catenin, enhances its nuclear localization and transcriptional activity, and occupies the b-catenin target vimentin promoter. All these hTERT effects were independent of its telomere-lengthening function or telomerase activity. hTERT and EMT marker expression correlates positively in GC samples. Mouse experiments demonstrate the in vivo stimulation of hTERT on cancer cell colonization. Collectively, hTERT stimulates EMT and induces stemness of cancer cells, thereby promoting cancer metastasis and recurrence. Thus, targeting hTERT may prevent cancer progression by inhibiting EMT and CSCs. Oncogene (2013) 32, 4203–4213; doi:10.1038/onc.2012.441; published online 8 October 2012 Keywords: CSCs; EMT; gastric cancer; hTERT

INTRODUCTION Metastasis is the most common cause of human cancer death, and the underlying mechanism remains elusive.1 Evidence has recently accumulated that epithelial–mesenchymal transition (EMT) has a critical role in cancer metastatic progression.2–6 EMT is a highly conserved developmental program that allows polarized, immotile epithelial cells to convert to those with motile mesenchymal properties. The EMT program is regulated by developmental transcriptional factors, including the zinc finger proteins Snail 1/2, and the basic helix–loop–helix factors Twist1/2 and Six, and the TGF-b and Wnt/b-catenin signaling pathways. These transcription factors or signaling pathways repress epithelial marker E-cadherin expression, whereas they induce the expression of mesenchymal vimentin and N-cadherin.2–7 The acquistion of mesenchymal features promote motility and invasiveness in malignant cells. Moreover, cancer cell EMT is associated with stemness, immunosuppression and resistance to treatment.8–10 All the above effects mediated by EMT significantly contribute to cancer progression. The aberrant expression of EMT transcription factors and/or the mesenchymal marker vimentin has been observed in various cancers, and generally predicts poor patients’ outcomes.11–15 Telomerase is a RNA-dependent DNA polymerase that elongates the 50 -TTAGGG-30 telomeric DNA.16 Most normal human

somatic cells lack telomerase activity because of the tight transcriptional repression of its rate-limiting, catalytic component telomerase reverse transcriptase (hTERT) gene.16,17 It is well established that stabilization of telomere length is required for transformed cells to achieve infinite proliferation potential during oncogenesis, and activation of telomerase/induction of hTERT expression to maintain telomere sizes is a critical step in transformation.17,18 hTERT expression and telomerase activation are observed in up to 90% of human malignancies,16,19 and targeting telomerase or telomere structure for cancer therapy has been suggested.20–22 In addition to its requirement for cancer development by maintaining telomere length, studies further indicate multiple functions of hTERT in oncogenesis.23–25 Ectopic expression of hTERT was previously shown to promote malignant transformation independently of telomere lengthening;26 telomerase and hTERT may protect cancer cells from apoptosis mediated by various insults;27–30 and higher hTERT expression is associated with advanced diseases and unfavorable prognosis in different types of malignancies.31–35 It is incompletely understood how telomerase or hTERT contributes to cancer progression. Given an important role of EMT in cancer invasion and metastasis, we sought to explore a link of hTERT with EMT using gastric cancer (GC) cells as a model.

1 Department of Microbiology/Key Laboratory for Experimental Teratology of Chinese Ministry of Education, School of Medicine and Second Hospital, Shandong University, Jinan, PR China.; 2Department of Medicine, Division of Haematology and Centre for Molecular Medicine (CMM), Karolinska University Hospital and Karolinska Institutet, Stockholm, Sweden; 3Department of Biosciences and Nutrition, Karolinska University Hospital and Karolinska Institutet, Stockholm, Sweden; 4Department of Women and Child Health, Karolinska University Hospital and Karolinska Institutet, Stockholm, Sweden; 5Laboratory of Cell Signaling and Carcinogenesis, Van Andel Research Institute, Grand Rapids, MI, USA and 6Central Research Laboratory, the Second Hospital of Shandong University, Jinan, PR China. Correspondence: Professor J Jia, Department of Microbiology/Key Laboratory for Experimental Teratology of Chinese Ministry of Education, School of Medicine and Second Hospital, Shandong University, Jinan 250012, PR China or Dr D Xu, Department of Medicine, Division of Haematology and Centre for Molecular Medicine (CMM), Karolinska University Hospital and Karolinska Institutet, L8:00 171 76 Stockholm, Sweden. E-mail: [email protected] or [email protected] Received 19 March 2012; revised and accepted 9 August 2012; published online 8 October 2012

EMT and CSC phenotype induced by hTERT Z Liu et al

4204 RESULTS hTERT overexpression induces a mesenchymal phenotype in GC cells and stimulates tumor cell invasion When studying the effect of hTERT overexpression on BGC-823 and KATOIII cells, we noticed that fibroblastoid-like alterations occurred in the subpopulation (up to 30%) of BGC-823 cells infected with a hTERT retroviral vector (hTERT-BGC-823)

(Figures 1a and b). We thus analyzed a panel of epithelial and mesenchymal markers in control and hTERT-BGC-823 cells. Basic levels of E-Cadherin, Snail1 and vimentin expression were detectable in control BGC-823 cells infected with an empty pBabe vector (pBabe-BGC-823) (Figure 1c). There was a slight reduction in E-Cadherin expression, whereas enhanced expression of vimentin and Snail1 was readily observed in hTERT-BGC-823 cells.

Figure 1. Overexpression of hTERT stimulates EMT-like alteration in morphology, Snail1 and vimentin expression, and invasion of GC cells. (a) EMT-like changes occurring morphologically in the subpopulation of hTERT-overexpressing BGC-823 cells. The cells were infected with pBabe-BGC-823 (left panel) or pBabe-hTERT-BGC-823 (right panel) retroviral vectors and examined under light microscopy after 1 week selection. (b) hTERT mRNA expression and telomerase activity in pBabe- and hTERT-BGC-823/KATOIII cells. The levels of hTERT mRNA (left panel) and telomerase activity (right panel) were assessed using a qPCR and Telomerase ELISA kit, respectively, and expressed in arbitrary units based on three independent experiments. Bars: s.d. (c) Immunoblotting analyses of hTERT, E-Cadherin, Snail1 and vimentin protein expression in pBabe- and hTERT-BGC-823/KATOIII cells. (d) Vimentin expression as determined using immunofluorescence assay in pBabe(upper panel) and hTERT-BGC-823 cells (lower panel). (e) Snail1 and vimentin mRNA expression in pBabe- and hTERT-BGC-823/KATOIII cells, as determined using qPCR. (f ) Matrigel assay for the invasion ability of pBabe- and hTERT-BGC-823/KATOIII cells. (g) The number of cells passing through a Matrigel filter. Bars: s.d. (h) HGC-27 GC cells were infected with the retroviral vector encoding wild-type hTERT, and then analyzed for Snail1 and vimentin protein levels (upper panel) and telomerase activity (lower panel) using immunoblotting and Telomerase ELISA kit, respectively. Shown is one representative of two independent experiments. Bars: s.d. Oncogene (2013) 4203 – 4213

& 2013 Macmillan Publishers Limited

EMT and CSC phenotype induced by hTERT Z Liu et al

4205 The increase in vimentin protein level was further confirmed by an immunofluorescence assay (Figure 1d). Moreover, Snail1 and vimentin mRNA expression was significantly upregulated in hTERT-BGC-823 cells compared with their control counterparts (Figure 1e), which indicates that hTERT overexpression stimulates transcription of both genes. Because the EMT program activation contributes to cancer cell invasion, we determined the invasive potential of hTERT-BGC-823 cells using Matrigel Matrix. DAPI staining revealed the cells passing through the Matrigel filter, and we observed a significantly increased invasiveness of hTERT-BGC823 cells compared with pBabe-BGC-823 cells (pBabe vs hTERT (mean±s.d. throughout the paper): 125±20 vs 186±25/1  105 cells, P ¼ 0.026, Figures 1f and g). Two more GC cell lines were further tested for a link between hTERT and EMT markers. KATOIII cells were infected with hTERT expression vectors and then analyzed for the EMT phenotype. Many of the pBabe- and hTERT-KATOIII cells are in suspension, and it is difficult to determine their morphological changes. Nevertheless, higher Snail1 and vimentin expression was similarly seen in hTERT-KATOIII cells compared with that in pBabe-KATOIII cells (Figures 1a–e). Furthermore, overexpression of hTERT drove significantly higher numbers of cells to pass through the Matrigel filter (pBabe vs hTERT: 85±7 vs 126±9/1  105 cells, Po0.001, Figures 1f and g). As seen in BGC-823 and KATOIII cells, induction of Snail1 and vimentin expression also occurred in third GC cell line HGC-27 expressing ectopic hTERT (Figure 1h). hTERT inhibition results in loss of mesenchymal markers and diminished invasion of tumor cells We next asked whether hTERT inhibition negatively affected the EMT program and cell invasion. We knocked down hTERT expression accompanied by diminished telomerase activity using hTERT-specific small interfering RNA (siRNA) (Figures 2a and b). Knocking down hTERT expression led to marked declines in the abundance of Snail1 and vimentin protein in both BGC-823 and KATOIII cells (Figure 2b). Consistent with their expression at protein levels, Snail1 and vimentin transcripts were significantly downregulated, as determined using quantitative PCR (qPCR) (Figure 2c). In addition, the downregulation of Snail1 and vimentin expression was similarly observed in HGC-27 cells treated with hTERT siRNA (Figure 2d). Fewer hTERT-depleted BGC-823 and KATOIII cells passed through the Matrigel filter compared with the control siRNA-treated cells (control vs hTERT: BGC-823: 124±47 vs 38±20/1  105 cells, P ¼ 0.008; KATOIII: 83±7 vs 47±6/1  105 cells, Po0.001; Figures 2e and f). hTERT enhances the number of GC stem cells (CSCs) and expression of CSC marker CD44 Because cancer cells having undergone EMT can acquire stem cell properties,8 we determined the presence of CSCs using the mono-sphere formation assay. As shown in Figure 3a, hTERT overexpression in BGC-823 cells enhanced the formation of monospheres by eightfold. Because KATOIII cells lacked the spheregeneration capacity under such culture conditions (data not shown), we were unable to make the same comparison between parent and hTERT-overexpressed KATOIII cells. We then analyzed the expression of CD44, an established gastric CSC marker,36 and a significant increase in CD44 expression was observed on hTERToverexpressed cells compared with their control counterparts (Figure 3b). In contrast, siRNA-mediated hTERT depletion led to a significant decline of CD44 levels in BGC-823 and KATOIII cells (Figure 3c). To further corroborate the inhibitory effect of hTERT depletion on cancer cell stemness, we sought to determine OCT-4, a well-known stem cell marker that is expressed in both normal cells and CSCs, and an OCT-4 promoter-driven green fluorescent protein (GFP) expression was previously used to label and identify those stem cells.37,38 We transfected BGC-823 cells with an OCT-4 & 2013 Macmillan Publishers Limited

promoter-driven GFP vector, and then determined GFP expression in the cells treated with hTERT siRNA. A significant reduction of GFP þ cells took place in hTERT-depleted BGC-823 cells, (Figure 3d) consistent with the alteration of CD44 expression profile found in these cells. hTERT inhibition and overexpression impairs and stimulates TGF-b1-mediated EMT, respectively TGF-b1 has an important role in activating EMT program in cancer cells through signaling pathways, including Smads, Snail/Slug and Wnt/b-catenin.5,39,40 We thus explored a potential involvement of hTERT in TGF-b1-mediated EMT. TGF-b1 treatment of KATOIII cells induced Snail1 and vimentin expression (Figures 4a and b). However, hTERT inhibition by siRNA substantially abolished TGF-b1-stimulated Snail1 and vimentin induction at both mRNA and protein levels (Figures 4a and b), whereas hTERT-overexpressing cells exhibited much higher levels of Snail1 and vimentin proteins than did control cells in response to TGF-b1 (Figure 4c), suggesting a synergistic effect of TGF-b1 and hTERT on EMT induction. hTERT affects TGF-b1-mediated b-catenin induction and nuclear accumulation The Wnt/b-catenin signaling pathway is one of the key factors activating the EMT program and participates in TGF-b1-induced EMT as well.5,41,42 Furthermore, recent studies showed that TERT regulated expression of a subset of b-catenin target genes.43 We thus probed whether b-catenin is the downstream effector affected by hTERT in the TGF-b1-mediated EMT. We first determined whether b-catenin is required for this TGF-b1 effect in our setting. As shown in Figure 4c, knocking down b-catenin resulted in diminished expression of Snail1 and vimentin. Moreover, TGF-b1-induced Snail1 and vimentin expression was significantly blocked in b-catenin-depleted KATOIII cells (Figure 4d). Taken together, b-catenin is involved in the TGF-b1mediated EMT. We thus further examined b-catenin expression and distribution. Both hTERT and TGF-b1 affected b-catenin protein expression: slight but consistent increase in hTERToverexpressed or TGF-b1-treated cells (Figures 4c–f), whereas slight reduction in hTERT-depleted cells (Figure 4g). The nuclear accumulation of b-catenin was evident in KATOIII cells treated with TGF-b1 (Figure 4e and Supplementary Figure S2). However, this TGF-b1 effect on b-catenin induction and nuclear distribution was markedly abolished when hTERT expression was inhibited in KATOIII cells (Figures 4e and f, and Supplementary Figure S2). In contrast, hTERT overexpression enhanced TGF-b1-mediated upregulation of b-catenin expression (Figure 4c). hTERT interacts with b-catenin and participates in the transcriptional regulation of downstream targets We further sought to explore potential interaction between b-catenin and hTERT using immunoprecipitation (IP). Because of lack of specific hTERT antibodies for IP, we employed an alternative strategy: 293T cells were transfected with flag-bcatenin, and GFP-hTERT expression vectors and cellular proteins were then analyzed using IP with antibodies against flag and GFP. Reciprocal IP showed a physical association of hTERT with b-catenin (Figure 4h). To corroborate this interaction, we then performed IP analyses on BGC-823 cells co-transfected with hTERT-HA and flag-b-catenin expression vectors. hTERT-HA and flag-b-catenin were highly expressed in those transfected cells, as determined using immunoblotting (Supplementary Figure S3A). As shown in Figure 4i, HA signals were detectable in the materials precipitated with a flag antibody, further demonstrating presence of the hTERT/b-catenin complex in a different setting. In all the above IP experiments, we used immunoglobulin G instead of GFP or flag antibodies to precipitate the same amounts of tagged proteins as negative controls. Oncogene (2013) 4203 – 4213

EMT and CSC phenotype induced by hTERT Z Liu et al

4206

Figure 2. Inhibition of hTERT expression leads to downregulation of Snail1 and vimentin expression and diminished invasion capacity of BGC-823 and KATOIII cells. (a) hTERT mRNA and telomerase activity using qPCR and telomerase ELISA kit, respectively. (b) Immunoblotting analyses of hTERT, E-Cadherin, Snail1 and vimentin protein expression in cells treated with control (C) and hTERT (h) siRNAs. (c) Snail1 and vimentin mRNA expression in cells treated with control (con) and hTERT siRNAs. (d) HGC-27 GC cells were transfected with hTERT siRNA and then analyzed for Snail1 and vimentin protein levels (upper panel) and telomerase activity (lower panel). Shown is one representative of two independent experiments. C and h, control and hTERT siRNA, respectively. (e) Matrigel assay for the invasion ability of BGC-823 and KATOIII cells treated with control and hTERT siRNAs. (f ) The number of cells passing through a Matrigel filter. Con, control.

To assess the functional consequence of the observed interaction, we determined the effect of hTERT on the TCF/LEF reporter activity. In hTERT-overexpressed KATOIII and BGC-823 cells, the luciferase activity driven by the TCF/LEF transcriptional response elements increased significantly (Figure 4j, left panel). In contrast, the TCF/LEF transactivation was markedly inhibited in hTERT-depleted KATOIII and BGC-823 cells (Figure 4j, left panel and Figure 5c). An hTERT-negative cell line, U2OS, was further utilized to assess a potential synergistic effect of hTERT and b-catenin on the TCF/LEF reporter activity. As expected, ectopic Oncogene (2013) 4203 – 4213

hTERT alone did not affect, whereas b-catenin transfection led to TCF/LEF reporter activation. When co-transfection of hTERT and b-catenin expression vectors were performed, substantially increased TCF/LEF reporter activity was observed (Figure 4j, right panel). Collectively, hTERT is required for the optimal TCF/LEF transactivation via interaction with b-catenin. hTERT is associated with the vimentin promoter b-catenin targets the vimentin gene for its transcription.5 To directly probe evidence that hTERT regulates vimentin & 2013 Macmillan Publishers Limited

EMT and CSC phenotype induced by hTERT Z Liu et al

4207

Figure 3. hTERT affects self-renewal and CD44 marker expression of gastric CSCs. (a) Mono-spheroid colony formation of pBabe- and hTERTBGC-823 cells. (b) Flow cytometry analyses of gastric CSC marker CD44 expression in pBabe- and hTERT-BGC-823/KATOIII cells. (c) Gastric CSC marker CD44 expression in BGC-823 and KATOIII cells treated with control and hTERT siRNAs. (d) OCT-4 promoter-driven GFP expression in BGC-823 cells treated with control and hTERT siRNAs. BGC-823 cells were transfected with a GFP vector driven by the OCT-4 promoter and selected for transfectants. These cells were then transfected with control and hTERT siRNA and analyzed for GFP-positive cells using flow cytometry.

transcription in cooperation with b-catenin, we determined their presence on the vimentin promoter using chromatin immunoprecipitation (ChIP) assay (Supplementary Figure S3B). Because there are no commercially available hTERT antibodies for the ChIP assay, we transfected 293T cells with flag-b-catenin and GFPhTERT expression vectors, and ChIP analysis was then performed using antibodies against flag and GFP. As shown in Figure 4k, an evident signal for the vimentin promoter was seen in the cells transfected with flag-b-catenin and GFP-hTERT plasmids using flag or GFP antibodies for precipitation. & 2013 Macmillan Publishers Limited

hTERT-stimulated EMT marker expression and b-catenin transactivation does not require its catalytic activity or telomerelengthening function To address whether the stimulatory effect of hTERT on EMT marker expression is dependent on its catalytic or telomere-lengthening function, we transfected BGC-823 cells with the vectors that express following hTERT variants: (i) the dominant-negative hTERT (mutation D869A) that loses a catalytic function. This hTERT variant inhibits telomerase activity by competing with the wildtype hTERT. (ii) C-terminal HA-tagged hTERT that exhibits Oncogene (2013) 4203 – 4213

EMT and CSC phenotype induced by hTERT Z Liu et al

4208 telomerase activity but is unable to elongate telomere in vivo.44 hTERT-HA, although exhibits telomerase activity in experimental tubes, perturbs its interaction with other telomere proteins, thereby preventing its access to and elongation of telomeres in living cells. Consistent with previous studies, DN-hTERT and hTERT-HA expression significantly inhibited and enhanced telomerase activity in BGC-823 cells, respectively (Figure 5a). Immunoblotting analyses revealed increased Snail1 and vimentin expression in BGC-823 cells transfected with DN-hTERT and hTERTHA vectors regardless of telomerase activity (Figures 5a and b).

Oncogene (2013) 4203 – 4213

Such effects of DN-hTERT and hTERT-HA were similarly observed in another GC cell line HGC-27 (Figure 5e). In addition, we assessed telomere length using both quantitative-fluorescent in situ hybridization and Flow-fluorescent in situ hybridization in hTERTdepleted BGC, where significant inhibition of Snail1 and vimentin expression occurred (Figures 2a–c), and their telomere was not shorter compared with the control cells (Supplementary Figure S1 and data not shown). The TCF/LEF reporter activity was decreased in hTERT-depleted BGC-823 cells, whereas enhanced in hTERT-overexpressed ones

& 2013 Macmillan Publishers Limited

EMT and CSC phenotype induced by hTERT Z Liu et al

4209 (Figure 5c). We thus further determined whether DN-hTERT and hTERT-HA, like wild-type hTERT, are capable of promoting the b-catenin-mediated target transactivation. To this end, hTERTnegative U2OS cells were transfected with the TCF/LEF reporter construct plus the vectors encoding b-catenin or/and DN-hTERT or/and hTERT-HA. Both DN-hTERT and hTERT-HA significantly enhanced b-catenin-mediated TCF/LEF reporter activation (Figure 5d).

hTERT promotes in vivo tumorigenesis and colonization of GC cells We then asked whether hTERT overexpression promotes in vivo metastasis of cancer cells. We thus injected pBabe-BGC-823 and hTERT-BGC-823 cells into nude mice through the tail vein and then examined their lungs for tumor seeding. Two doses of both pBabe-BGC-823 and hTERT-BGC-823 cells (6  105 and 2  105/ mouse, respectively) were injected (8–10 mice per subgroup), and mice were killed after three weeks. Both pBabe-BGC-823 and

Figure 5. hTERT regulation of EMT marker expression and b-catenin transcriptional activity is independent on its catalytic function. (a) BGC823 cells were transfected with DN-hTERT and hTERT-HA expression vectors and then analyzed for Snail1 and vimentin expression at 72 h using immunoblotting. (b) Telomerase activity in the same sets of BGC-823 cells was determined using a telomerase ELISA kit. Bars: s.d. (c) Regulation of TCF/LEF reporter activity by hTERT. hTERT-depleted or overexpressed BGC-823 cells were transfected with TCF/LEF reporters and luciferase activity was assessed using the dual reporter assay kit. Bars: s.d. Con: control. (d) hTERT-deficient U2OS cells were transfected with TCF/LEF reporters together with b-catenin vectors alone or b-catenin plus DN-hTERT or hTERT-HA expression vectors and luciferase activity was assessed 48 h post-transfection. Bars: s.d. (e) HGC-27 cells were transfected with expression vectors encoding DN-hTERT and hTERT-HA and then analyzed for Snail and vimentin expression using immunoblotting. Telomerase activity was determined using a telomerase ELISA kit and and expressed in arbitrary units (lower panels). Shown is one representative of two independent experiments.

Figure 4. TGF-b1 and b-catenin pathways are involved in hTERT-mediated EMT. (a) and (b) Attenuation of TGF-b1-mediated EMT by hTERT depletion in KATOIII cells. Snail1 and vimentin mRNA and protein expression was analyzed using qPCR and western blot, respectively. (c) A synergistic effect of TGF-b1and hTERT on b-catenin, Snail1 and vimentin induction. pBabe-KATOIII and hTERT-KATOIII cells were incubated with TGF-b1 overnight, and then analyzed for transcripts and/or protein levels of each gene using qPCR (right panel) and immunoblotting (left panel), respectively. (d) Diminished Snail1 and vimentin protein expression and attenuated TGF-b1-mediated EMT by b-catenin inhibition. Left panel: the levels of Snail1 and vimentin protein were reduced in b-catenin siRNA-treated KATOIII cells. Right panel: TGF-b1-stimulated Snail1 and vimentin expression was blocked by knocking down b-catenin. (e) Abolishment of TGF-b1-induced nuclear accumulation of b-catenin by hTERT depletion. KATOIII cells were transfected with control or hTERT siRNA for 48 h and then incubated with TGF-b1 overnight. The distribution of b-catenin was analyzed using immunofluorescence (IF) assay. (f ) The fluorescence intensity of b-catenin signals in (e) was quantified using NIS software (Nikon Instech Co., Ltd, Tokyo, Japan) (mean±s.d. per cell) and expressed in arbitrary units. (g) b-catenin expression in hTERT-depleted KATOIII cells, as determined using immunoblotting. Shown is one representative of two independent experiments. (h) Physical interaction between hTERT and b-catenin. 293T cells were transfected with GFP-hTERT or GFP and flag-b-catenin expression vectors and a reciprocal IP was performed using GFP or flag antibodies. (i) BGC-823 cells were co-transfected with flag-b-catenin and hTERT-HA or control vectors and IP was performed using flag antibody. The presence of hTERT-HA in the precipitated materials was detected using HA antibody. (j) Regulation of TCF/LEF reporter activity by hTERT. Left panel: hTERT depletion and overexpression inhibits and stimulates TCF/LEF transactivation, respectively. Right panel: hTERT-deficient U2OS cells were transfected with TCF/LEF reporters together with b-catenin vectors alone or b-catenin plus hTERT expression vectors, and luciferase activity was assessed 48 h post-transfection. (k) ChIP assay for b-catenin and hTERT occupancy on the vimentin promoter. 293T cells were co-transfected with GFP-hTERT and flag-b-catenin expression vectors and ChIP was performed with chromatin derived from these cells. The signal enrichment with each antibody is quantified using qPCR. Shown is one representative of two independent experiments. & 2013 Macmillan Publishers Limited

Oncogene (2013) 4203 – 4213

EMT and CSC phenotype induced by hTERT Z Liu et al

4210

Figure 6. hTERT overexpression promotes in vivo colonization of BGC-823 cells. (a) Lungs derived from non-injected, pBabe- and hTERT-BGC823 injected mice. (b) Lung weight (mean±s.d., n ¼ 8–10) between pBabe- and hTERT-BGC-823 mouse groups. (c) hTERT mRNA levels in tumors derived from pBabe- and hTERT-BGC-823 mice. (d) Hematoxylin and eosin staining of lung tissues.

Figure 7. hTERT mRNA expression is associated with Snail1 and vimentin mRNA levels in primary GC tissues. (a) and (b) Tumors derived from 31 patients were examined for hTERT, Snail1 and vimentin mRNA expression using qPCR and the levels were expressed in arbitrary units. A linear relationship between hTERT and Snail1 or vimentin mRNA was evaluated with a regression analysis. (c) Higher levels of hTERT, Snail1 and vimentin mRNA expression in gastric tumors 43 cm (in diameter). Cancer samples were divided into two groups (p3 cm and 43 cm) according to their volumes (in diameter), and the gene expression was compared between these two groups. Bars: s.d.

hTERT-BGC-823 cells formed metastastic colonies, and the lungs from mice injected with hTERT-BGC-823 cells were significantly bigger and heavier than those from pBabe-BGC-823-bearing mice (control vs hTERT in lung weights (g): 0.90±0.08 vs 1.58±0.2, 95% confidence interval: 0.5–1.1 vs 1.2–1.8, Po0.001] (Figures 6a and b). More tumor colonies were observed in the lungs from mice harboring hTERT-BGC-823 cells than pBabe-BGC-823 cells (pBabe-BGC-823 vs hTERT-BGC-823 cells: 32±7 per lung area vs 54±10 per lung area, P ¼ 0.034) (Figure 6d). Tumors derived from hTERT-BGC-823 cells expressed much higher levels of hTERT mRNA compared with those from pBabe-BGC-823 cells (Figure 6c). hTERT expression is associated with Snail1 and vimentin in primary GC Finally, we determined hTERT, Snail1 and vimentin expression in 31 primary tumor samples from GC patients using qPCR. hTERT, Snail1 and vimentin mRNA was expressed in 25/31 (81%), 17/31 (55%) and 10/31 (32%) of cancer samples, respectively. A highly significant correlation between hTERT and Snail1 or vimentin Oncogene (2013) 4203 – 4213

transcripts was observed in these samples (Figures 7a and b). Furthermore, tumors 43.0 cm (in diameter) exhibited significantly higher levels of three transcripts than did smaller ones (p3.0 cm) (Figure 7c). Strikingly, vimentin mRNA was found only in one of 14 tumors of p3 cm.

DISCUSSION It has been well established that induction of hTERT expression and telomerase activation is prerequisite to cellular immortalization and malignant transformation.18 However, whether and how hTERT is involved in cancer progression remains incompletely understood. In the present study, we provide evidence that hTERT has a significant role in regulating EMT and stemness of GC cells, thereby contributing to cancer progression. In clinical studies, higher levels of hTERT expression are found to be associated with poor outcomes of various human malignancies.31–35 Experimentally, cells overexpressing hTERT are more resistant to different insults, including chemotherapeutic & 2013 Macmillan Publishers Limited

EMT and CSC phenotype induced by hTERT Z Liu et al

4211 treatment.27,28 hTERT-mediated EMT provides one putative explanation for these observations. Indeed, EMT activation not only stimulates epithelial cancer invasion and metastasis through making motile cancer cells, but also contributes to resistance to chemotherapy by regulating various events.8–10,45,46 More recently, EMT has been shown to be associated with acquisition of CSC properties,6,8–10 while CSCs are required for the formation of metastases from disseminated tumor cells.47,48 The direct link between EMT and stemness likely promotes more invasive behavior of cancer cells. These activities of EMT are at least partially consistent with the observed role of hTERT in cancer progression. Recent studies unveil multiple biological activities of telomerase or hTERT in addition to its telomere-lengthening function. TERT overexpression enhanced normal stem cell mobilization and proliferation independently of telomerase activity.49 Mechanistically, TERT achieves this by stimulating the transcription of b-catenin target genes.43 Our findings suggest that the effect of hTERT on EMT and stemness of GC cells is associated with b-catenin signaling activation too. First, hTERT overexpression and depletion enhanced and inhibited TCF/LEF transcriptional activity, respectively. Second, hTERT and b-catenin synergized to activate the TCF/LEF response elements. Finally, the presence of the hTERT and b-catenin complex was demonstrated. Because hTERT is nucleus-localized, its physical interaction with b-catenin may prevent b-catenin degradation and promote b-catenin nuclear retention. Consistently, we did observe that hTERT overexpression and depletion led to up- and downregulation of b-catenin expression, respectively. Moreover, TGF-b1-mediated nuclear accumulation of b-catenin was abolished by inhibiting hTERT expression. Collectively, our result reveals a novel mechanism through which hTERT regulates the transcription of b-catenin target genes and EMT program, and furthermore, these effects of hTERT are independent on its catalytic activity. Although both Park et al.43 and we demonstrated that TERT is required for optimal activation of b-catenin target genes, the mechanism underlying TERT action is different. Park et al.43 found that mTERT enhances the transcription of b-catenin target genes through interacting with Brg1, which is an important factor in the chromatin-remodeling complex SWI/SNF. However, we failed to see a physical association between hTERT and Brg1 in human cells (data not shown). Instead, our results demonstrate that hTERT directly interacts with b-catenin, and thereby prevents its degradation and nuclear retention. One of putative explanations for the discrepancy is that two different mechanisms are employed by human and mouse TERT, respectively, to regulate b-catenin target genes. b-catenin is not only involved in TGF-b1-mediated EMT, but also directly promotes EMT phenotype.41,50,51 Moreover, the EMTstimulatory factor, Snail1, interacts with b-catenin and enhances the transcription of its target genes.52 Such a positive feedback loop is believed to amplify the EMT effect. Our present findings demonstrate that hTERT serves as an important partner in EMT mediated by b-catenin. Moreover, hTERT promotes the transactivation of b-catenin target genes independently of its catalytic activity. However, mTERT  /  cells have recently been shown to have wild-type levels of Wnt signaling in vitro,53 which is in contrast to the findings by Park et al.43 It is currently unclear why there exist such discrepancies among different reports, and putative explanations can be differences in species, cell types, TERT expression levels (overexpression or physiological level) and experimental settings in those different studies. We show that hTERT stimulates the EMT program through the Wnt/b-catenin pathway. In addition, hTERT overexpression dramatically increased the number of spheroid colonies in BGC-823 cells, which suggests an important role for hTERT in CSC selfrenewal. Consistently, it has recently been shown that CSCs exhibit abundant amounts of telomerase activity, and disruption of & 2013 Macmillan Publishers Limited

telomere maintenance suppresses the self-renewal of CSCs.54 Taken together, targeting hTERT may help eliminate CSCs, thereby preventing cancer progression. We believe that future clinical trials designed to evaluate the efficacy of telomerase inhibitors on the prevention of cancer metastasis, and recurrence will be of importance. MATERIALS AND METHODS Cell Lines, reagents, siRNA transfection and retroviral vector infection Human GC cell lines BGC-823 (purchased from Cancer Institute Beijing, Beijing, China, 2008) and KATOIII cells (purchased from CLS-Cell Line Service, Eppelheim, Germany, 2008) free of mycoplasma (Checked using PCR, 2009) were used in the present study. TGF-b1 was from Sigma-Aldrich (St Louis, MO, USA). We bought chemical-modified Stealth siRNAs from Invitrogen (Carlsbad, CA, USA). pBabe and pBabe-hTERT retroviral vectors were used to infect BGC-823 and KATOIII cells.

Mono-sphere formation assay Cells were cultured in ultra-low-attachment 96-well plates (Corning Life Sciences, Amsterdam, The Netherlands) with 100 ml RPMI-1640/10 mM HEPES serum-free medium supplemented with cocktails of following growth factors: human recombinant EGF 20 ng/ml, and human recombinant bFGF (Invitrogen) 10 ng/ml. After 10 days, each well was examined using light microscope and the number of spheroid colonies was counted.

RNA extraction, reverse transcription and qPCR Total cellular RNA was extracted using ULTRASPEC kit (Biotecx Lab, Houston, TX, USA). cDNA was synthesized and qPCR was carried out in an ABI7700 sequence detector (Applied Biosystems, Foster City, CA, USA) with SYBR Green kit (Applied Biosystems). Levels of target mRNA were calculated based on the CT values and normalization of human b2-M expression.

Assessment of telomerase activity Telomerase activity was assayed with a commercial Telomerase PCR enzyme-linked immunosorbent assay (ELISA) kit (Roche Diagnostics, Scandinavia AB, Sweden). For each assay, 1 mg of protein was used, and 26 PCR cycles were performed. The PCR products were detected using ELISA color reaction.

Flow cytometry Cells were stained with CD44-PE (BD Biosciences, Sparks, MD, USA) and the corresponding isotype control (Becton Dickinson, Franklin Lakes, NJ, USA). After incubation and washing, CD44 expression was detected with a flow cytometer (Becton Dickinson); OCT-4-driven GFP-expressing BGC-823 cells were determined as above without antibody staining. The results were analyzed using CellQuest software (Becton Dickinson).

Western blot and IP Total cellular proteins were extracted with RIPA lysis buffer. For IP, 200 mg proteins were precipitated with 2 mg of polyclonal antibodies against GFP (Santa Cruz Biotechnology, Santa Cruz, CA, USA) or monoclonal antibodies against flag (Sigma-Aldrich). Two micrograms of rabbit or mouse immunoglobulin G (Sigma-Aldrich) served as an isotype control in IP experiments. The precipitated material, or 30 mg of proteins in case of western blot analyses, was subjected to SDS–polyacrylamide gel electrophoresis and transferred to a nitrocellulose membrane. The membranes were probed with the specific antibodies against hTERT (Rockland Immunochemicals, San Diego, CA, USA), E-Cadherin, Snail1, vimentin (Cell Signaling Technology, Billerica, MA, USA), b-catenin (BD Biosciences), HA (Santa Cruz Biotechnology), flag or b-actin (Sigma-Aldrich) and then with secondary anti-mouse or rabbit horseradish peroxidase-conjugated immunoglobulin G (Bio-Rad, Hercules, CA, USA).

Matrigel invasion assay The Matrigel invasion membrane (BD BioCoat Matrigel Invasion Chamber; BD Biosciences) was rehydrated. Cells were added to a Matrigel invasion chamber in triplicate and incubated for 8 h. After noninvading cells were removed, the membranes containing the invading cells were stained with DAPI. Oncogene (2013) 4203 – 4213

EMT and CSC phenotype induced by hTERT Z Liu et al

4212 TCF/LEF reporter activity assay

REFERENCES

The TCF/LEF reporter kit was from SABiosciences (Valencia, CA, USA). To determine the effect of hTERT on the TCF/LEF reporter activity, we transfected hTERT-overexpressed or -depleted cells with TCF/LEF reporter plasmids. TCF/LEF-driven luciferase activity was determined using a dual luciferase reporter assay system (Promega, Madison, WI, USA) 48 h posttransfection, and the target promoter-driven firefly luciferase activity was normalized to the renilla activity included in the kit.

1 Nguyen DX, Bos PD, Massague J. Metastasis: from dissemination to organ-specific colonization. Nat Rev Cancer 2009; 9: 274–284. 2 Kalluri R, Weinberg RA. The basics of epithelial-mesenchymal transition. J Clin Invest 2009; 119: 1420–1428. 3 Polyak K, Weinberg RA. Transitions between epithelial and mesenchymal states: acquisition of malignant and stem cell traits. Nat Rev Cancer 2009; 9: 265–273. 4 Tsuji T, Ibaragi S, Hu GF. Epithelial-mesenchymal transition and cell cooperativity in metastasis. Cancer Res 2009; 69: 7135–7139. 5 van Zijl F, Zulehner G, Petz M, Schneller D, Kornauth C, Hau M et al. Epithelial-mesenchymal transition in hepatocellular carcinoma. Future Oncol 2009; 5: 1169–1179. 6 Thiery JP, Acloque H, Huang RY, Nieto MA. Epithelial-mesenchymal transitions in development and disease. Cell 2009; 139: 871–890. 7 Zeisberg M, Neilson EG. Biomarkers for epithelial-mesenchymal transitions. J Clin Invest 2009; 119: 1429–1437. 8 Mani SA, Guo W, Liao MJ, Eaton EN, Ayyanan A, Zhou AY et al. The epithelialmesenchymal transition generates cells with properties of stem cells. Cell 2008; 133: 704–715. 9 Kurrey NK, Jalgaonkar SP, Joglekar AV, Ghanate AD, Chaskar PD, Doiphode RY et al. Snail and slug mediate radioresistance and chemoresistance by antagonizing p53-mediated apoptosis and acquiring a stem-like phenotype in ovarian cancer cells. Stem Cells 2009; 27: 2059–2068. 10 Gupta PB, Onder TT, Jiang G, Tao K, Kuperwasser C, Weinberg RA et al. Identification of selective inhibitors of cancer stem cells by high-throughput screening. Cell 2009; 138: 645–659. 11 Creighton CJ, Li X, Landis M, Dixon JM, Neumeister VM, Sjolund A et al. Residual breast cancers after conventional therapy display mesenchymal as well as tumor-initiating features. Proc Natl Acad Sci USA 2009; 106: 13820–13825. 12 Waldmann J, Feldmann G, Slater EP, Langer P, Buchholz M, Ramaswamy A et al. Expression of the zinc-finger transcription factor Snail in adrenocortical carcinoma is associated with decreased survival. Br J Cancer 2008; 99: 1900–1907. 13 Kim MA, Lee HS, Lee HE, Kim JH, Yang HK, Kim WH. Prognostic importance of epithelial-mesenchymal transition-related protein expression in gastric carcinoma. Histopathology 2009; 54: 442–451. 14 Hayry V, Salmenkivi K, Arola J, Heikkila P, Haglund C, Sariola H. High frequency of SNAIL-expressing cells confirms and predicts metastatic potential of phaeochromocytoma. Endocr Relat Cancer 2009; 16: 1211–1218. 15 Rosivatz E, Becker I, Specht K, Fricke E, Luber B, Busch R et al. Differential expression of the epithelial-mesenchymal transition regulators snail, SIP1, and twist in gastric cancer. Am J Pathol 2002; 161: 1881–1891. 16 Liu JP, Chen SM, Cong YS, Nicholls C, Zhou SF, Tao ZZ et al. Regulation of telomerase activity by apparently opposing elements. Ageing Res Rev 2010; 9: 245–256. 17 Shay JW, Wright WE. Hallmarks of telomeres in ageing research. J Pathol 2007; 211: 114–123. 18 Hanahan D, Weinberg RA. The hallmarks of cancer. Cell 2000; 100: 57–70. 19 Shay JW, Bacchetti S. A survey of telomerase activity in human cancer. Eur J Cancer 1997; 33: 787–791. 20 Chen H, Li Y, Tollefsbol TO. Strategies targeting telomerase inhibition. Mol Biotechnol 2009; 41: 194–199. 21 Harley CB. Telomerase and cancer therapeutics. Nat Rev Cancer 2008; 8: 167–179. 22 Boukamp P, Mirancea N. Telomeres rather than telomerase a key target for anti-cancer therapy? Exp Dermatol 2007; 16: 71–79. 23 Zhou L, Zheng D, Wang M, Cong YS. Telomerase reverse transcriptase activates the expression of vascular endothelial growth factor independent of telomerase activity. Biochem Biophys Res Commun 2009; 386: 739–743. 24 Cong Y, Shay JW. Actions of human telomerase beyond telomeres. Cell Res 2008; 18: 725–732. 25 Hamad NM, Banik SS, Counter CM. Mutational analysis defines a minimum level of telomerase activity required for tumourigenic growth of human cells. Oncogene 2002; 21: 7121–7125. 26 Stewart SA, Hahn WC, O’Connor BF, Banner E, Lundberg A, Modha P et al. Telomerase contributes to tumorigenesis by a telomere length-independent mechanism. Proc Natl Acad Sci USA 2002; 99: 12606–12611. 27 Dudognon C, Pendino F, Hillion J, Saumet A, Lanotte M, Segal-Bendirdjian E. Death receptor signaling regulatory function for telomerase: hTERT abolishes TRAIL-induced apoptosis, independently of telomere maintenance. Oncogene 2004; 23: 7469–7474. 28 Luiten RM, Pene J, Yssel H, Spits H. Ectopic hTERT expression extends the life span of human CD4 þ helper and regulatory T-cell clones and confers resistance to oxidative stress-induced apoptosis. Blood 2003; 101: 4512–4519. 29 Zhang X, Mar V, Zhou W, Harrington L, Robinson MO. Telomere shortening and apoptosis in telomerase-inhibited human tumor cells. Genes Dev 1999; 13: 2388–2399.

Chromatin immunoprecipitation ChIP assay was done as described previously.55 293T cells transfected with different plasmids were cross-linked by incubating them in 1% (vol/vol) formaldehyde-containing medium for 10 mins at 37 1C and then sonicated to make soluble chromatin. The antibodies against flag (Sigma-Aldrich) and GFP (Santa Cruz Biotechnology) were used to precipitate DNA fragments bound by their corresponding elements. The protein–DNA complex was collected with protein A Sepharose beads (Millipore, Billerica, MA, USA), eluted, and reverse cross-linked. Following treatment with Protease K (Sigma-Aldrich), the samples were extracted with phenol–chloroform and precipitated with ethanol. The recovered DNA was re-suspended in TE buffer and quantified using qPCR with vimentin promoter and GAPDH primers listed in Supplementary Table S1.

Immunofluorescence Cells were treated with TGF-b and/or hTERT siRNA. Primary antibodies against vimentin or b-catenin were added at a dilution of 1:500 for 30 min at room temperature after blocking steps with background buster. The cells were then incubated with a Cy3-conjugated donkey anti-rabbit antibody (1:600; Jackson Immunoresearch, West Grove, PA, USA). The nuclei were counterstained with DAPI. The results were analyzed with a Leica TCS SP5 microscope (Leica Mikrosysteme Vertrieb GmbH, Wetzlar, Germany) with filters for the detection of DAPI and Cy3.

Patients Thirty-one patients with primary GC who underwent gastrectomy at Qilu Hospital, Shandong University in 2007 were included in the study. The tumor specimens were collected after surgery and stored at  80 1C until use. Patients’ information including age, gender, tumor, histology, differentiation status, size (diameter), invasiveness, and regional and distant metastases at the time of operation was in Supplementary Table S2. The study was approved by the local ethics committees.

Mouse model pBabe-BGC-823 and hTERT-BGC-823 cells were collected, and 6  105 and 2  105 cells were suspended in 100 ml phosphate-buffered saline, respectively. The cells were then injected into nude mice (Shanghai Slac Laboratory Animal Co. Ltd, Shanghai, China) via the tail vein. Mice were killed after 3 weeks and lungs were collected for evaluation of tumor seeding or metastasis.

Statistical analyses Mann—Whitney U-test or Student’s t-test was used for analyses of differences between experiment groups. Correlation analyses of hTERT, Snail1 and vimentin in GC samples were made using linear regression. All the tests were two-tailed and computed using SigmaStat3.1 software (Systat Software Inc., Richmond, CA, USA). P-values of o0.05 were considered as statistically significant.

CONFLICT OF INTEREST The authors declare no conflict of interest.

ACKNOWLEDGEMENTS We thank Drs RA Weinberg (Massachusetts Institute of Technology), CM Counter (Duke University), JM Wong and K Collins (University of California, Berkeley), Y Cong (Beijing Normal University) and W Cui (Imperial College London) for retroviral vectors and plasmids.This study was supported by the National Basic Research Program of China (grant No. 973 Program 2012CB911202), the Swedish Cancer Society, the Swedish Research Council, Cancer Society in Stockholm, Swedish Child Cancer Society, the Karolinska Institutet Foundations, National Natural Science Foundation of China (No.: 30770118, 30800406, 30972775, 81071721, 81000868, 81171536), the National Key Scientific Program of China (2007CB914801) and the Science Foundation of Shandong Province (No.: ZR2009CZ001, ZR2009CM002 and BS2010YY040)

Oncogene (2013) 4203 – 4213

& 2013 Macmillan Publishers Limited

EMT and CSC phenotype induced by hTERT Z Liu et al

4213 30 Cao Y, Li H, Deb S, Liu JP. TERT regulates cell survival independent of telomerase enzymatic activity. Oncogene 2002; 21: 3130–3138. 31 Domont J, Pawlik TM, Boige V, Rose M, Weber JC, Hoff PM et al. Catalytic subunit of human telomerase reverse transcriptase is an independent predictor of survival in patients undergoing curative resection of hepatic colorectal metastases: a multicenter analysis. J Clin Oncol 2005; 23: 3086–3093. 32 Gertler R, Rosenberg R, Stricker D, Friederichs J, Hoos A, Werner M et al. Telomere length and human telomerase reverse transcriptase expression as markers for progression and prognosis of colorectal carcinoma. J Clin Oncol 2004; 22: 1807–1814. 33 Terrin L, Rampazzo E, Pucciarelli S, Agostini M, Bertorelle R, Esposito G et al. Relationship between tumor and plasma levels of hTERT mRNA in patients with colorectal cancer: implications for monitoring of neoplastic disease. Clin Cancer Res 2008; 14: 7444–7451. 34 Tabori U, Ma J, Carter M, Zielenska M, Rutka J, Bouffet E et al. Human telomere reverse transcriptase expression predicts progression and survival in pediatric intracranial ependymoma. J Clin Oncol 2006; 24: 1522–1528. 35 Ohali A, Avigad S, Cohen IJ, Meller I, Kollender Y, Issakov J et al. Association between telomerase activity and outcome in patients with nonmetastatic Ewing family of tumors. J Clin Oncol 2003; 21: 3836–3843. 36 Takaishi S, Okumura T, Tu S, Wang SS, Shibata W, Vigneshwaran R et al. Identification of gastric cancer stem cells using the cell surface marker CD44. Stem Cells 2009; 27: 1006–1020. 37 Levings PP, McGarry SV, Currie TP, Nickerson DM, McClellan S, Ghivizzani SC et al. Expression of an exogenous human Oct-4 promoter identifies tumor-initiating cells in osteosarcoma. Cancer Res 2009; 69: 5648–5655. 38 Gerrard L, Zhao D, Clark AJ, Cui W. Stably transfected human embryonic stem cell clones express OCT4-specific green fluorescent protein and maintain self-renewal and pluripotency. Stem Cells 2005; 23: 124–133. 39 Vincent T, Neve EP, Johnson JR, Kukalev A, Rojo F, Albanell J et al. A SNAIL1SMAD3/4 transcriptional repressor complex promotes TGF-beta mediated epithelial-mesenchymal transition. Nat Cell Biol 2009; 11: 943–950. 40 Chaudhury A, Hussey GS, Ray PS, Jin G, Fox PL, Howe PH. TGF-beta-mediated phosphorylation of hnRNP E1 induces EMT via transcript-selective translational induction of Dab2 and ILEI. Nat Cell Biol 2010; 12: 286–293. 41 Zhou BP, Deng J, Xia W, Xu J, Li YM, Gunduz M et al. Dual regulation of Snail by GSK-3beta-mediated phosphorylation in control of epithelial-mesenchymal transition. Nat Cell Biol 2004; 6: 931–940. 42 Medici D, Hay ED, Olsen BR. Snail and Slug promote epithelial-mesenchymal transition through beta-catenin-T-cell factor-4-dependent expression of transforming growth factor-beta3. Mol Biol Cell 2008; 19: 4875–4887.

43 Park JI, Venteicher AS, Hong JY, Choi J, Jun S, Shkreli M et al. Telomerase modulates Wnt signalling by association with target gene chromatin. Nature 2009; 460: 66–72. 44 Counter CM, Hahn WC, Wei W, Caddle SD, Beijersbergen RL, Lansdorp PM et al. Dissociation among in vitro telomerase activity, telomere maintenance, and cellular immortalization. Proc Natl Acad Sci USA 1998; 95: 14723–14728. 45 Fuchs BC, Fujii T, Dorfman JD, Goodwin JM, Zhu AX, Lanuti M et al. Epithelialto-mesenchymal transition and integrin-linked kinase mediate sensitivity to epidermal growth factor receptor inhibition in human hepatoma cells. Cancer Res 2008; 68: 2391–2399. 46 Kudo-Saito C, Shirako H, Takeuchi T, Kawakami Y. Cancer metastasis is accelerated through immunosuppression during Snail-induced EMT of cancer cells. Cancer Cell 2009; 15: 195–206. 47 Visvader JE, Lindeman GJ. Cancer stem cells in solid tumours: accumulating evidence and unresolved questions. Nat Rev Cancer 2008; 8: 755–768. 48 Liu H, Patel MR, Prescher JA, Patsialou A, Qian D, Lin J et al. Cancer stem cells from human breast tumors are involved in spontaneous metastases in orthotopic mouse models. Proc Natl Acad Sci USA 2010; 107: 18115–18120. 49 Sarin KY, Cheung P, Gilison D, Lee E, Tennen RI, Wang E et al. Conditional telomerase induction causes proliferation of hair follicle stem cells. Nature 2005; 436: 1048–1052. 50 Dissanayake SK, Wade M, Johnson CE, O’Connell MP, Leotlela PD, French AD et al. The Wnt5A/protein kinase C pathway mediates motility in melanoma cells via the inhibition of metastasis suppressors and initiation of an epithelial to mesenchymal transition. J Biol Chem 2007; 282: 17259–17271. 51 Yook JI, Li XY, Ota I, Hu C, Kim HS, Cha SY et al. A Wnt-Axin2-GSK3beta cascade regulates Snail1 activity in breast cancer cells. Nat Cell Biol 2006; 8: 1398–1406. 52 Stemmer V, de Craene B, Berx G, Behrens J. Snail promotes Wnt target gene expression and interacts with beta-catenin. Oncogene 2008; 27: 5075–5080. 53 Strong MA, Vidal-Cardenas SL, Karim B, Yu H, Guo N, Greider CW. Phenotypes in mTERT/ and mTERT/ mice are due to short telomeres, not telomere-independent functions of telomerase reverse transcriptase. Mol Cell Biol 2011; 31: 2369–2379. 54 Marian CO, Cho SK, McEllin BM, Maher EA, Hatanpaa KJ, Madden CJ et al. The telomerase antagonist, imetelstat, efficiently targets glioblastoma tumor-initiating cells leading to decreased proliferation and tumor growth. Clin Cancer Res 2010; 16: 154–163. 55 Liu L, Liu C, Lou F, Zhang G, Wang X, Fan Y et al. Activation of telomerase by seminal plasma in malignant and normal cervical epithelial cells. J Pathol 2011; 225: 203–211.

Supplementary Information accompanies the paper on the Oncogene website (http://www.nature.com/onc)

& 2013 Macmillan Publishers Limited

Oncogene (2013) 4203 – 4213