CCAAT/Enhancer binding protein β induces motility and invasion of glioblastoma cells through transcriptional regulation of the calcium binding protein S100A4

Oncotarget4369 www.impactjournals.com/oncotarget www.impactjournals.com/oncotarget/ Oncotarget, Vol. 6, No.6 CCAAT/Enhancer binding protein β induces motility and invasion of glioblastoma cells through transcriptional regulation of the calcium binding protein S100A4 Diana Aguilar-Morante1,2,4, Jose A. Morales-Garcia1, Angel Santos2,3, Ana Perez- Castillo1,2 1 Instituto de Investigaciones Biomédicas, (CSIC-UAM). Departamento Modelos Experimentales de Enfermedades Humanas, Arturo Duperier, 4. 28029-Madrid, Spain 2Centro de Investigación Biomédica en Red sobre Enfermedades Neurodegenerativas (CIBERNED), Spain 3 Departamento de Bioquímica y Biología Molecular, Facultad de Medicina, Universidad Complutense de Madrid, 28040-Madrid, Spain 4 Instituto de Biomedicina de Sevilla, IBiS, (Hospital Universitario Virgen del Rocío/CSIC/Universidad de Sevilla). Departamento de Fisiología Médica y Biofísica, 41013-Sevilla, Spain Correspondence to: Ana Perez-Castillo, e-mail: [email protected] Angel Santos, e-mail: [email protected] Keywords: C/EBPβ, S100A4, invasiveness, transcriptional activity, cancer stem cells Received: August 13, 2014 Accepted: December 20, 2014 Published: January 24, 2015 ABSTRACT We have previously shown that decreased expression of CCAAT/Enhancer binding protein β (C/EBPβ) inhibits the growth of glioblastoma cells and diminishes their transformation capacity and migration. In agreement with this, we showed that C/ EBPβ depletion decreases the mRNA levels of different genes involved in metastasis and invasion. Among these, we found S100 calcium binding protein A4 (S100A4) to be almost undetectable in glioblastoma cells deficient in C/EBPβ. Here, we have evaluated the possible role of S100A4 in the observed effects of C/EBPβ in glioblastoma cells and the mechanism through which S100A4 levels are controlled by C/EBPβ. Our results show that C/EBPβ suppression significantly reduced the levels of S100A4 in murine GL261 and human T98G glioblastoma cells. By employing an S100A4-promoter reporter, we observed a significant induction in the transcriptional activation of the S100A4 gene by C/EBPβ. Furthermore, overexpression of S100A4 in C/EBPβdepleted glioblastoma cells reverses the enhanced migration and motility induced by this transcription factor. Our data also point to a role of S100A4 in glioblastoma cell invasion and suggest that the C/EBPβ gene controls the invasive potential of GL261 and T98G cells through direct regulation of S100A4. Finally, this study indicates a role of C/EBPβ on the maintenance of the stem cell population present in GL261 glioblastoma cells. INTRODUCTION Glioblastomas (GBM) constitute approximately 50% of gliomas as well as 20% of intracranial tumors. These tumors are characterized by uncontrolled cell proliferation, necrosis focus, high angiogenic activity, apoptosis resistance and diffuse invasion of neoplastic cells into the surrounding tissue, making surgical interventions rather ineffective [1]. GBMs represent the most malignant type of astrocytomas and, despite advances in treatment, they remain refractory to conventional therapies and the medium survival of patients is only one year after diagnosis [2]. This poor prognosis is probably due, at least in part, to the infiltrative nature of these tumors and the presence of cellular populations with ability to escape therapies and drive tumor recurrence and progression. In some cases, these resistant cells exhibit stem properties (Glioblastoma Stem cells (GSC) [3]. CCAAT/Enhancer Binding Protein β (C/EBPβ), is a member of a family of transcription factors consisting of Oncotarget4370 www.impactjournals.com/oncotarget six structurally related basic leucine-zipper DNA-binding protein [4]. C/EBPβ has important roles in numerous tissues regulating cellular proliferation and differentiation, metabolism, adipogenesis, inflammation, transformation and tumorigenesis [5, 6]. This protein is expressed in the central nervous system where it plays different roles [7, 8]. Our group has shown that C/EBPβ is an important factor for neuronal differentiation [9] and regulates the expression of several genes involved in inflammatory processes and brain injury [10]. Besides, mice lacking C/EBPβ show a reduced inflammatory response after an excitotoxic insult and are less susceptible to neuronal cell loss [11]. The effects of C/EBPβ in tumor development is controversial; some evidence suggests that C/EBPβ acts as potent promoter of tumorigenesis while others indicate that it can have antiproliferative effects. For instance, ectopic expression of C/EBPβ in primary fibroblast causes cell cycle arrest and is required for RasV12-induced senescence [12]. By contrast, it has been observed that C/EBPβ is abundant in colorectal tumors [13] and ovarian cancer [14] and its expression is associated with tumor progression. Recently, some data show that C/EBPβ is associated with tumor progression in prostate cancer cells and regulates the expression of metastatic genes in these tumors [15, 16]. Regarding brain tumors, Homma et al. (2006) has shown that expression of C/EBPβ is increased in high grade gliomas compared with less aggressive gliomas, and patients where the expression of C/EBPβ is lower have a longer survival [17]. In this regard, we have demonstrated that C/EBPβ plays an important role in the regulation of proliferation, migration and invasiveness of glioblastoma cells [18]. Furthermore, our results showed that a decrease in C/EBPβ expression is associated with reduced mRNA levels of different genes involved in invasiveness and metastasis, including S100A4 [18]. S100A4, also known as mts1, is a member of the S100 family of Ca2+-binding proteins and is thought to be directly involved in tumor invasion and metastasis via interactions with specific protein targets [19] in a variety of tumors, such as breast cancer [20, 21], gastric cancer [22], pancreatic cancer [23], colorectal cancer [24], nonsmall cell lung cancer [25], and prostate cancer [26]. Recent data show that expression of S100A4 in tumors of the central nervous system is related to the degree of malignancy of the tumor, with a higher expression of S100A4 in high-grade gliomas compared to low-grade gliomas [27–29]. All these data together prompted us to analyze whether C/EBPβ is a direct regulator of S100A4 expression and if this protein could mediate the observed effects of C/EBPβ on migration and invasiveness of glioblastoma cells. Also, we have studied the effects of C/ EBPβ depletion on the neurosphere formation of mouse glioblastoma GL261cells. Collectively, these studies show that C/EBPβ increases S100A4 levels by directly activating S100A4 promoter expression in glioblastoma cells and provide evidence that S100A4 may contribute to the C/EBPβinduced invasiveness. Also, we demonstrate that S100A4 promotes migration and invasiveness of glioblastoma cells. Finally, our results indicate that C/EBPβ depletion resulted in an inhibition of proliferation and self-renewal of these cells, which is accompanied by a decreased in the S100A4 protein levels. RESULTS Regulation of S100A4 gene expression by C/EBPβ transcription factor In a previous work [18] we screened two arrays of cell cycle and cancer-related genes in order to analyze the mechanism involved in the tumorigenic effect of C/EBPβ in glioblastoma cells. We identified a number of genes involved in different processes, such as cell cycle regulation, DNA damage response, adhesion, invasion, and metastasis to be differentially expressed in those cells deficient in C/ EBPβ. Among these genes, S100A4 showed extremely low levels of expression in the I4 cells. Here we have deeply analyzed the regulation of this gene by C/EBPβ. Figure 1A shows that the levels of S100A4 mRNA, as measured by RT-PCR were significantly decreased (16- and 15-fold) in those cells deficient in C/EBPβ (pools I4 and I5, respectively) compared to controls (pool C1). In addition to this reduction in the mRNA levels of S100A4, S100A4 protein levels were almost undetectable in these cells (Figure 1B). These data are consistent with the PCR arrays data suggesting a possible link between C/EBPβ and S100A4 during the progression of the tumorigenic capacity of glioblastoma cells. We next tested whether other members of the S100 family, also involved in tumorigenic processes, were also regulated by C/EBPβ. As can be seen in Figure 1C, the expression of S100A6, A100A8, and S100A10 was also down-regulated in those glioblastoma cells deficient in C/EBPβ (pool I4). C/EBPβ is a direct transcriptional regulator of S100A4 To further analyze the role of C/EBPβ in regulating S100A4 gene expression, we studied whether C/EBPβ regulates S100A4 promoter activity. First, we performed an in silico analysis to search for putative C/EBPβ binding sites in the S100A4 promoter using TFSEARCH and MatInspector programs. We identified a consensus binding site for this transcription factor at the position –606/–591 (cut-off value: 0.95) suggesting that C/EBPβ may directly regulate S100A4 expression. We also found a putative binding region for the transcription Oncotarget4371 www.impactjournals.com/oncotarget factor AP1 at the –680/–670 position (cut-off value: 0.84) (Figure 2A). It has been described that there are some interactions between C/EBPβ and AP1 in the regulation of gene expression. C/EBPβ can bind to AP1 binding elements as homodimers and activate the transcription of the target gene, whereas its heterodimerization with Fos or Jun leads to an alteration of the DNA binding specificity of C/EBPβ to C/EBPβ DNA binding sites [30] To provide direct evidence that C/EBPβ is recruited to the endogenous S100A4 promoter during transcription in vivo, we performed standard chromatin immunoprecipitation (ChIP) assays, which allow the detection of proteins bound to specific regions of DNA in vivo. For these assays we used GL261 cells and a specific antibody against C/EBPβ to precipitate the complex formed by DNA and this transcription factor. ChIP analysis with C/EBPβ antibody showed binding of this transcription factor to the S100A4 promoter at the C/EBPβ and AP1 consensus regions (Figure 2B). As a control of specificity of the assay, binding of C/EBPβ to the housekeeping β-actin locus was not detected. These data indicate that C/EBPβ can interact with two regions of the S100A4 promoter. We next performed transient transfection experiments to determine whether C/EBPβ regulates the activity of the S100A4 gene by cotransfecting the S100A4 promoter-luciferase construct together with the pcDNA3/C/EBPβ overexpression construct in GL261 cells (Figure 2C). GL261 cells cotransfected with pcDNA3/ C/EBPβ plasmid displayed a significant increase in the promoter activity of S100A4. Deletion of both C/EBPβ and AP1 binding sites resulted in a significantly reduced C/EBPβ-induced activity. Subsequently, to assess the functional role of the C/EBPβ and AP1 binding sites in S100A4 promoter regulation, we performed site-specific mutagenesis within these binding regions. We transfected the mutated luciferase reporters into GL261 cells and compared their activity with that of wild-type S100A4 promoter. As shown in Figure 2D, disruption of C/EBPβ or both C/EBPβ and AP1 binding sites significantly attenuated Figure1: Effect of C/EBPβ on S100A4 expression in glioblastoma murine GL261 cell line. (A) Quantification of S100A4 mRNA levels in GL261 control cell line (C1) and C/EBPβ-depleted (I4, I5) cells by quantitative real time-PCR. As indicated in Methods, we used Taqman probes specifics to S100A4 and β-Actin mouse. The graphic shown the means of values 2−∆∆Ct of S100A4/β-actin ± SD. ***p < 0.001. (B) Representative Western blot showing expression of C/EBPβ and S100A4 in C1, I4 and I5 cell lines. (C) Quantification of S100A6, S100A8 and S100A10 mRNA levels in GL261 control cell line (C1) and C/EBPβ-depleted (I4) cells by quantitative real time-PCR. As indicated in Methods, we used Fast SYBR Green and primers specifics to S100A6, S100A8, S100A10 and mouse β-Actin. The graphic shown the means of values 2−∆∆Ct ± SD. ***p < 0.001. Oncotarget4372 www.impactjournals.com/oncotarget Figure 2: C/EBPβ actives S100A4 promoter. (A) Schematic diagram of S100A4 promoter showing the localization of C/EBPβ and AP1 binding sites. (B) Representative image and quantification of ChIP analysis of C/EBPβ binding to the endogenous S100A4 promoter in GL261 cells. DNA before immunoprecipitation was used as positive control (Input) and a region in the actin gene was used as a negative control. Data are expressed relative to the input values and represent the mean ± SD determined in three independent experiments. (C) and (D) Transient transfection experiments. The entire promoter fragment (pS100A4/1248), a 5’-deletion construct (pS100A4/298), and two constructs containing the mutated C/EBPβ (pS100A4/Mut-C/EBPβ) and the mutated C/EBPβ and AP1 binding sites (pS100A4//Mut- C/EBPβ-AP1) were created as indicated in Methods. Data are expressed relative to the basal values and represent the mean ± SD luciferase activity determined in triplicate in at least three independent experiments. ***p < 0.001. Oncotarget4373 www.impactjournals.com/oncotarget S100A4 promoter activity although this reduction was less evident than that observed with the deletion construct. S100A4 overexpression in C/EBPβ-depleted cell line reversed invasion and motility induced by C/EBPβ Glioblastoma cells are characterized for their capacity to invade normal surrounding tissue. S100A4 is a well-known inductor of tumor cell motility and metastasis in different cancer cells [31], although its role in the invasiveness capacity of glioblastoma cells is not yet known. Previous data from our laboratory, using the “scratch-wound” assay showed that GL261 glioblastoma cells depleted of C/EBPβ presented a restricted cell motility [18]. Here we have analyzed the effect of S100A4 overexpression (Figure 3A) in invasion (transwell assay) and motility (scratch assay) of control and C/EBPβdepleted GL261glioblastoma cells. As shown in Figure 3B, overexpression of S100A4 caused a clear increased in the invasion capacity of C1, I4 and I5, being more marked in the C/EBPβ-depleted cells. The I4 and I5 cells transfected with the pIRES2-DsRed-Express vector containing S100A4 cDNA showed an increase in their invasion capacity to levels similar to those found in the non-depleted cells (Figure 3B). Regarding cell motility, our results clearly Figure 3: Effect of C/EBPβ and S100A4 on GL261 cells invasion capacity. (A) Representative Western blots showing S100A4 levels in all the cell lines used for the invasion analysis. (B) The invasion capacity of C1, I4 and I5 cells transfected with the S100A4 expressing vector pIRES2-DsRed-Express or the corresponding control vector were determined on transwell chambers coated with Collagen Type IV as described in Methods. Representative images and quantifications are presented. Values are the means ± S.D. of three different experiments. Scale bar 50 μm. **p < 0.01; *p < 0.05. Oncotarget4374 www.impactjournals.com/oncotarget show that the overexpression of S100A4 increased motility in all GL261 cells, C1, I4 and I5 (Figure 4A and 4B). These results are in accordance with the invasion results described above and suggest that S100A4 gene controls the motility of glioblastoma cells and that therefore could mediate the effects of C/EBPβ on motility and invasion. Next, we determined the effect of C/EBPβ depletion and S100A4 overexpression (Figure 5A and 5B) on invasion and motility in the human glioblastoma cell line T98G. Similar to the result observed in GL261 cells, C/ EBPβ depletion in T98G cells caused a marked decrease in S100A4 protein content (Figure 5A), invasion capacity (Figure 5C and 5D) and cell motility (Figure 6). The overexpression of S100A4 in T98G cells, as in GL261 cells, increased invasion capacity (Figure 5C) and motility (Figure 6A and 6B) in both non C/EBPβ-depleted (TC) and C/EBPβ-depleted (TI) cells. Since it has been shown [32, 33] that S100A4 regulates the expression of invasion- and migrationassociated genes, such as metalloproteases (MMPs), we next analyzed the mRNA levels of MMP2, which is known to be regulated by S100A4 in esophageal squamous cell carcinoma [34]. Indeed, we found a significant decrease in MMP2 expression in the C/EBPβ-depleted I4 cells (Supplemental Figure S1). These results further confirm the idea that S100A4 is a mediator of the effects of C/EBPβ in the invasion and motility capacities of glioblastoma cells. In order to know whether S100A4 could also play a role in proliferation of GL261 and T98G cells, we performed “QIA127 Rapid Cell Proliferation Kit” assay both in control and C/EBPβ-depleted cells when S100A4 is overexpressed. As can be shown in Figure 7, we didn’t find any difference in growth and viability between control and S100A4- overexpressing cells. These results are in agreement with data from Takenaga et al [35] showing that S100A4 is not involved in the growth of C6 glioblastoma cells. Effect of C/EBPβ depletion on glioblastoma stem cells The cancer stem cell (CSC) hypothesis suggests that tumors are organized in a hierarchy with a subpopulation of cells with stem cells properties, responsible for tumor maintenance and progression. Studies from our laboratory indicate that C/EBPβ is implicated in neural stem cells proliferation and differentiation in normal adult mouse brain [8]. In view of these data and based on our previous results showing that C/EBPβ plays an important role in the progression of tumorigenicity of glioblastoma cells, we finally analyzed whether C/EBPβ interference could also exert an antiproliferative effect on glioblastoma stem cells by analyzing its effect in glioblastoma-derived neurospheres (GNSF) and also if S100A4 levels were down-regulated in C/EBPβ-depleted GNSFs. To ensure that we were working with an enriched population in GNSF, we first analyzed by western blot the protein levels of Musashi-1 and Oct3/4, which are wellknown stem cells markers [36]. As expected, adherent C1 and I4 cells did not have Mushashi-1 and Oct3/4 proteins, while GNSF populations of PC1 and PI4 presented detectable levels of these markers (Figure 8A). Also, we observed a decrease in the levels of Musashi-1 in PI4 cells, compared to PC1, suggesting that C/EBPβ can alter the population of stem cells present in these cultures. Next, we analyzed the basal level of expression of C/EBPβ in murine GNSF populations obtained from control C1 line and C/EBPβ-depleted I4 cells (PC1 and PI4 pools, respectively). Our results show that as happened in adherent C1 and I4 cells, PI4 cultures presented a significant (80%) decrease in the expression levels of C/EBPβ (Figure 8B). This reduction in C/EBPβ levels was accompanied by a large reduction in S100A4 levels, indicating that C/EBPβ is also able to regulate S100A4 expression in the subpopulation of glioblastoma stem cells. We then studied whether C/EBPβ interference could affect the formation capacity and self-renewal of GNSF population in GL261 glioblastoma cells. Figure 8C shows that C/EBPβ depletion decreases the formation of primary glioblastoma neurospheres. We observed a significant decrease, at 7 days of culture, in the number of PI4 glioblastoma neurospheres, as compared to control PC1 cultures. To analyze more deeply the effect of C/ EBPβ depletion on the ability of GBM cells to generate new spheres, actively growing 7 day-old GL261-derived primary GNSF cultures were dissociated, and equal numbers of viable cells were replated in fresh neurosphere medium to generate new GNSF (secondary glioblastomaderived neurospheres). After 14 days, we observed that the formation of secondary GNSF were significantly less in the C/EBPβ-depleted cells (PI4), compared with control, PC1 (63 ± 21.5 and 235 ± 46.5 per plate, respectively) (Figure 8D). Furthermore, to test for the effect of C/ EBPβ interference on the self-renewal of the glioblastoma neurospheres, we dissociated established 7 days-old primary GNSF cultures and plated them at very low density during 7 and 14 days. As shown in Figure 8E, 8C/EBPβdepleted cultures (PI4) presented less number of secondary GNSF, 7 and 14 days after plating, compared with control population, PC1 (14 days after plating values were 8 ± 2.5 and 19 ± 3.0 per plate, respectively), indicating that these cultures contain less self-renewing stem cells. Lastly, we evaluated whether the S100A4 promoter was also regulated in the GNSF cultures by C/EBPβ. Consistent with the transfection data using GL261 adherent cells, we found a substantial increase in the basal activity of the pS100A4/1248 construct containing S100A4 promoter in those GNSF co-transfected with the pcDNA3/C/EBPβ overexpression construct (Figure 8F). Also, deletion of C/EBPβ and AP1 binding sites, as well as site-specific mutations in both sites, significantly attenuated C/EBPβ-induced activity of this promoter. Oncotarget4375 www.impactjournals.com/oncotarget Figure 4: Effect of C/EBPβ and S100A4 on GL261 cells motility. C1, I4 and I5 cells transfected with the S100A4 expression vector pIRES2-DsRed-Express or the corresponding control vector were grown until reach confluence. A linear scratch was performed with a plastic pipette tip. Images were taken with a phase contrast microscope at different times after wounding. (A) Representative phasecontrast images and (B) quantifications of the in vitro wound-healing assay are shown. Bar scale 100 μm. ***p < 0.001; *p < 0.05. Oncotarget4376 www.impactjournals.com/oncotarget DISCUSION Glioblastomas (GBM) are the most common brain tumor in adults. Currently this devastating disease is practically incurable and the patients show a mean survival time of approximately one year after diagnosis [37], despite the use of surgery, radiotherapy and chemotherapy [38]. Studies of gene expression in glioblastomas have allowed analyzing the transcriptional activity present in these tumors, allowing the classification and prediction of response to particular treatments [39–41]. In a previous work [18] we have demonstrated that C/EBPβ is crucial to regulate glioblastoma cell growth and transformation. These findings are in accordance with the results of Carro et al [42] showing that the expression of C/EBPβ is linked to the mesenchymal state of primary glioblastoma and provides an excellent prognostic biomarker for tumor aggressiveness. These authors showed that C/EBPβ, together with STAT3, act synergistically to initiate and regulate mesenchymal transformation. Overexpression of these transcription factors reprograms neural stem cells throughout the aberrant mesenchymal lineage and enhances tumor aggressiveness. Figure 5: Effect of C/EBPβ and S100A4 on T98G cells invasion capacity. (A) Representative western blot of C/EBPβ and S100A4 in control (TC) and C/EBPβ interfered (TI) T98G cells. (B) Representative western blot showing S100A4 levels in all the cell lines used for the invasion analysis. (C) The invasion capacity of TC and TI cells transfected with the S100A4 expressing vector pIRES2-DsRed- Express or the corresponding control vector was determined on transwell chambers coated with Collagen Type IV as described in Materials and Methods. (C) Representative images and (D) quantifications are presented. Values are the means ± S.D. of three different experiments. Scale bar 50 μm. ***p < 0.001; *p < 0.05. Oncotarget4377 www.impactjournals.com/oncotarget Also we showed that C/EBPβ regulates several genes involved in DNA repair, invasion and metastasis, suggesting that these genes could be important downstream effectors of C/EBPβ-mediated oncogenic properties [18]. One of these genes is S100A4, which is known to promote invasion and metastasis in different tumor cells [31]. In this regard, Yonemura et al. demonstrated that 55% of gastric cancer patients exhibited elevated S100A4 levels, which were found to be positively associated with high incidence of metastasis. This study also showed that patients with low expression of S100A4 have lower number of metastatic lesions [43]. With respect to glial tumors, it has been shown that S100A4 is expressed differentially in astrocytic tumors being its levels higher on those tumors with a higher degree of malignancy [44]. In the present study, we show that the transcription factor C/EBPβ directly regulates the expression of metastatic protein S100A4 in a mouse glioblastoma cell Figure 6: Effect of C/EBPβ and S100A4 on T98G cells motility. TC and TI cells transfected with the S100A4 expressing vector pIRES2-DsRed-Express or the corresponding control vector were grown until reach confluence. A linear scratch was performed with a plastic pipette tip. Images were taken with a phase contrast microscope at different times after wounding. (A) Representative images and (B) quantifications of the in vitro wound-healing assay are shown. Bar scale 100 μm. ***p < 0.001; **p < 0.01. Oncotarget4384 www.impactjournals.com/oncotarget 39. Brennan C, Momota H, Hambardzumyan D, Ozawa T, Tandon A, Pedraza A, Holland E. Glioblastoma subclasses can be defined by activity among signal transduction pathways and associated genomic alterations. PLoS One. 2009; 4:e7752. 40. Van Meir EG, Hadjipanayis CG, Norden AD, Shu HK, Wen PY, Olson JJ. Exciting new advances in neurooncology: the avenue to a cure for malignant glioma. CA Cancer J Clin. 2010; 60:166–193. 41. Verhaak RG, Hoadley KA, Purdom E, Wang V, Qi Y, Wilkerson MD, Miller CR, Ding L, Golub T, Mesirov JP, Alexe G, Lawrence M, O’Kelly M, Tamayo P, Weir BA, Gabriel S, et al. Integrated genomic analysis identifies clinically relevant subtypes of glioblastoma characterized by abnormalities in PDGFRA, IDH1, EGFR, and NF1. Cancer Cell. 2010; 17:98–110. 42. Carro MS, Lim WK, Alvarez MJ, Bollo RJ, Zhao X, Snyder EY, Sulman EP, Anne SL, Doetsch F, Colman H, Lasorella A, Aldape K, Califano A, Iavarone A. The transcriptional network for mesenchymal transformation of brain tumours. Nature. 2010; 463:318–325. 43. Yonemura Y, Endou Y, Kimura K, Fushida S, Bandou E, Taniguchi K, Kinoshita K, Ninomiya I, Sugiyama K, Heizmann CW, Schafer BW, Sasaki T. Inverse expression of S100A4 and E-cadherin is associated with metastatic potential in gastric cancer. Clin Cancer Res. 2000; 6:4234–4242. 44. Camby I, Nagy N, Lopes MB, Schafer BW, Maurage CA, Ruchoux MM, Murmann P, Pochet R, Heizmann CW, Brotchi J, Salmon I, Kiss R, Decaestecker C. Supratentorial pilocytic astrocytomas, astrocytomas, anaplastic astrocytomas and glioblastomas are characterized by a differential expression of S100 proteins. Brain Pathol. 1999; 9:1–19. 45. Gao XN, Tang SQ, Zhang XF. S100A4 antisense oligodeoxynucleotide suppresses invasive potential of neuroblastoma cells. J Pediatr Surg. 2005; 40:648–652. 46. Hernan R, Fasheh R, Calabrese C, Frank AJ, Maclean KH, Allard D, Barraclough R, Gilbertson RJ. ERBB2 up-regulates S100A4 and several other prometastatic genes in medulloblastoma. Cancer Res. 2003; 63:140–148. 47. Donato R, Cannon BR, Sorci G, Riuzzi F, Hsu K, Weber DJ, Geczy CL. Functions of S100 proteins. Curr Mol Med. 2013; 13:24–57. 48. Chen H, Xu C, Jin Q, Liu Z. S100 protein family in human cancer. Am J Cancer Res. 2014; 4:89–115. 49. Gross S, Sin C, Barraclough R, Rudland P. Joining S100 proteins and migration: for better or for worse, in sickness and in health. Cell Mol Life Sci. 2014; 71:1551–1579. 50. Kozlova EN, Lukanidin E. Mts1 protein expression in the central nervous system after injury. Glia. 2002; 37: 337–348. 51. Palm T, Schwamborn JC. Brain tumor stem cells. Biological Chemistry. 2010; 391:607–617. 52. Hambardzumyan D, Squatrito M, Carbajal E, Holland EC. Glioma formation, cancer stem cells, and akt signaling. Stem Cell Rev. 2008; 4:203–210. 53. Sanai N, Alvarez-Buylla A, Berger MS. Neural stem cells and the origin of gliomas. N Engl J Med. 2005; 353:811–822. 54. Singh SK, Clarke ID, Terasaki M, Bonn VE, Hawkins C, Squire J, Dirks PB. Identification of a cancer stem cell in human brain tumors. Cancer Res. 2003; 63:5821–5828. 55. Lo JF, Yu CC, Chiou SH, Huang CY, Jan CI, Lin SC, Liu CJ, Hu WY, Yu YH. The epithelial-mesenchymal transition mediator S100A4 maintains cancer-initiating cells in head and neck cancers. Cancer Res. 2011; 71:1912–1923. 56. Yan XL, Jia YL, Chen L, Zeng Q, Zhou JN, Fu CJ, Chen HX, Yuan HF, Li ZW, Shi L, Xu YC, Wang JX, Zhang XM, He LJ, Zhai C, Yue W, et al. Hepatocellular carcinoma-associated mesenchymal stem cells promote hepatocarcinoma progression: role of the S100A4-miR155- SOCS1-MMP9 axis. Hepatology. 2013; 57:2274–2286. 57. El Andaloussi A, Sonabend AM, Han Y, Lesniak MS. Stimulation of TLR9 with CpG ODN enhances apoptosis of glioma and prolongs the survival of mice with experimental brain tumors. Glia. 2006; 54:526–535. 58. Gine E, Morales-Garcia JA, Perez-Castillo A, Santos A. Developmental hypothyroidism increases the expression of kainate receptors in the hippocampus and the sensitivity to kainic acid-induced seizures in the rat. Endocrinology. 2010; 151:3267–3276. 59. Gomez-Cabello D, Callejas S, Benguria A, Moreno A, Alonso J, Palmero I. Regulation of the microRNA processor DGCR8 by the tumor suppressor ING1. Cancer Res. 2010; 70:1866–1874.