Oncogene (2012) 31, 1354–1365 & 2012 Macmillan Publishers Limited All rights reserved 0950-9232/12 www.nature.com/onc ORIGINAL ARTICLE Sox2 expression in breast tumours and activation in breast cancer stem cells

O Leis1, A Eguiara1, E Lopez-Arribillaga1, MJ Alberdi2, S Hernandez-Garcia3, K Elorriaga2, A Pandiella3, R Rezola2 and AG Martin1

1Regulation of Cell Growth Laboratory, Fundacio´n Inbiomed, San Sebastian, Spain; 2Department of Pathology, Onkologikoa, San Sebastian, Spain and 3Instituto de Biologia Molecular y Celular del Cancer, Salamanca, Spain

The cancer stem cell (CSC) model does not imply that of normal tissue stem cells based on germ-line variation, tumours are generated from transformed tissue stem cells. epigenetic change and somatic mutation of stem cell The target of transformation could be a tissue stem cell, a signalling components, and then acquire more malig- progenitor cell, or a differentiated cell that acquires self- nant phenotype based on accumulation of additional renewal ability. The observation that induced pluripotency epigenetic and genetic alterations, and tumour-stromal reprogramming and cancer are related has lead to the interaction at the invasion front (Katoh, 2010). How- speculation that CSCs may arise through a reprogram- ever, the target cell for transformation that originates ming-like mechanism. Expression of pluripotency genes CSCs remains an unknown issue. In solid tumours the (Oct4, Nanog and Sox2) was tested in breast tumours by phenotypic definition of CSCs is rather loose with no immunohistochemistry and it was found that Sox2 is single marker that unequivocally identifies that popula- expressed in early stage breast tumours. However, tion. Such poor definition has prevented so far the expression of Oct4 or Nanog was not found. Mammo- needed study of clinical relevance of CSC content in sphere formation in culture was used to reveal stem cell tumour management. There are a number of reports properties, where expression of Sox2, but not Oct4 or using in vitro culture of tumour cells and animal models Nanog, was induced. Over-expression of Sox2 increased showing that CSCs are more resistant to conventional mammosphere formation, effect dependent on continuous cancer therapies, therefore, placing these cells at the root Sox2 expression; furthermore, Sox2 knockdown pre- of tumour recurrence and metastases. Several prelimin- vented mammosphere formation and delayed tumour ary reports have indeed shown that this is the case formation in xenograft tumour initiation models. Induc- with human cancer patients. In breast cancer, Li et al. tion of Sox2 expression was achieved through activation (2008) showed that conventional chemotherapy in- of the distal enhancer of Sox2 promoter upon sphere creased the fraction of CD44 þ CD24À cells in a formation, the same element that controls Sox2 transcrip- neoadjuvant setting of advanced breast cancer patients tion in pluripotent stem cells. These findings suggest that and increased mammosphere formation in vitro. Tanei reactivation of Sox2 represents an early step in breast et al. (2009) have shown that paclitaxel and epirubicin- tumour initiation, explaining tumour heterogeneity by based chemotherapy enriches for aldehyde dehydrogen- placing the tumour-initiating event in any cell along the ase-1-positive cells in breast tumours, another marker axis of mammary differentiation. for CSCs (Ginestier et al., 2007). In an attempt to Oncogene (2012) 31, 1354–1365; doi:10.1038/onc.2011.338; associate CSC content and prognosis, the presence of published online 8 August 2011 CD44 þ CD24À cells in clinical samples of breast cancer was not associated with aggressive tumour behaviour Keywords: Sox2; breast cancer; cancer stem cell; and poor clinical outcome (Abraham et al., 2005), mammosphere indicating that the definition of CSCs must be first clearly established before it can have clinical use. It is possible that different breast cancer subtypes rely on specific CSC populations and, therefore, more precise Introduction functional identification must be developed. Reprogramming of adult terminally differentiated Many solid tumours, including breast cancer, exhibit a human cells into pluripotent stem cells capable of functional hierarchy of cells of which only a small generating all cells in the body has attracted a great subpopulation of stem-like cells give rise to the differ- deal of attention. The production of these induced entiated cells that make the bulk of the tumour (Al-Hajj pluripotent stem cells requires the over-expression of et al., 2003). Cancer stem cells (CSCs) occur as mimetic four transcription factors Oct4, Sox2, Klf4 and c-Myc (Takahashi et al., 2007), although Klf4 and c-Myc can Correspondence: Dr AG Martin, Regulation of Cell Growth be replaced by Lin28 and Nanog, and may even be Laboratory, Fundacio´n Inbiomed, Paseo Mikeletegi 61, San Sebastian, dispensable (Meissner et al., 2007; Nakagawa et al., Gipuzkoa 20009, Spain. E-mail: [email protected] 2008; Okita et al., 2008; Kim et al., 2009). The efficiency Received 17 January 2011; revised 16 June 2011; accepted 3 July 2011; of this reprogramming process is extremely low and published online 8 August 2011 remains so far as an in vitro phenomenon as there Sox2 activation in breast cancer stem cells O Leis et al 1355 is no evidence that it can naturally occur in vivo. sion of some or all of these genes has been found in The mechanisms underlying the reprogramming process several forms of human cancer. Using immunohisto- are not well understood yet, however, the three main chemistry we checked expression of these genes in a transcription factors Oct4, Sox2 and Nanog, called panel of human breast tumour samples representative of master regulators of pluripotency, have probed respon- the main breast cancer subtypes (Figure 1a). We sible for maintaining the undifferentiated state (Yama- included staining for b-catenin as activation of the naka, 2008). Recently, the process of reprogramming canonical Wnt signalling pathway, measured by nuclear and tumourigenesis (Kawamura et al., 2009; Li et al., localization of b-catenin, has been linked to mammary 2009; Marion et al., 2009; Utikal et al., 2009) have been organogenesis (Zeng and Nusse, 2010). For positive linked as the p53 tumour suppressor, one of the main controls, we used sections of seminoma (for Oct4 and regulators of oncogenic transformation, controls the Nanog), non-small cell lung carcinoma (for Sox2) and induction of pluripotency. colorectal carcinoma (for nuclear b-catenin). We did not Sox genes encode a family of high-mobility group observe expression of Oct4 or Nanog in any sample of transcription factors that have critical roles in organo- our series (n ¼ 81 for Oct4 and n ¼ 71 for Nanog), nor genesis. The functional specificity of different Sox nuclear b-catenin localization (n ¼ 64). However, we did proteins and the tissue specificity of a particular Sox observe Sox2-positive staining in B16% of the samples factor are largely determined by the differential partner- observed (n ¼ 158; Figure 1b1). Sox2 expression was ship with other transcription regulators, many of which previously reported in basal type breast cancer (Rodri- have not yet been discovered. Recently, a number of guez-Pinilla et al., 2007); in our tumour series we observed links have been found between the Sox2 and human Sox2-positive staining in tumour samples of all the cancers. Sox2 has been found to be an immunogenic major breast cancer subtypes (Table 1). For a complete antigen in 41% of small cell lung cancer patients (Gure description of tumour features see Supplementary et al., 2000) and in 29% of meningioma patients File 1. When tumours were classified according to their (Comtesse et al., 2005). Immunohistochemistry results TNM stage, Sox2-positive tumours fell into the early suggest that Sox2 is involved in invasion and metastasis stages of tumour progression: 57.9% in stage I, 26.3% in of pancreatic intraepithelial neoplasia (Sanada et al., stage IIA, 10.5% in stage IIB, 5.3% in stage IIIA and no 2006), and may also be involved in gastric carcinogen- Sox2-positive tumours in more advanced stages (IIIB or esis (Li et al., 2004) and prostate cancers (Sattler et al., beyond; Figure 1b2). Out of the 25 Sox2-positive tumours 2000). Furthermore, Sox2 expression has been observed found, 13 showed infiltrating tissue structure by histolo- in 43% of basal cell-like breast carcinomas, suggesting a gical examination, while 12 tumours still retained in situ role in conferring a less differentiated phenotype intra-ductal tumour component (Figure 1c). Interestingly, (Rodriguez-Pinilla et al., 2007). How Sox2 exerts its those tumours that retained intra-ductal tumour compo- oncogenic potential is currently unknown. nent showed significantly (P ¼ 0.0347 by paired t-test) In this work we investigated the relationship between higher numbers of Sox2-positive cells there than in the pluripotency induction and breast cancer by examining infiltrating part, suggesting the presence of Sox2-positive the presence of pluripotency-related genes in human cells in primitive areas of the tumour. Collectively, these breast tumours. We did not observe expression of Oct4 data indicate that Sox2 may be expressed in the initial or Nanog but we found expression of Sox2, which in stages of tumour formation and it is lost as the tumour turn was associated with early stages of tumour progresses towards advanced stages. formation. We observed induction of Sox2 expression in tumour spheres from natural breast tumour cultures and breast carcinoma cell lines, suggesting a role in Sox2 is expressed in tumour spheres tumour stem cells. This role was further investigated by Owing to the relevant role of Sox2 in pluripotency over-expressing and downregulating Sox2 expression, maintenance, we reasoned that Sox2 expression might and it was found that Sox2 was necessary and sufficient be related to the presence of CSCs. We utilized the to induce tumour sphere formation and tumour initia- tumour sphere assay as a surrogate in vitro culture assay tion in vivo. Regulation of Sox2 promoter activation for tumour formation to investigate Sox2 expression in probed to be dependent on the upstream enhancer and CSCs. In these culture conditions, CSCs are able to not on the core promoter, the same element that maintain stem-like properties such as tumour initiation, controls Sox2 expression in pluripotent stem cells. self-renewal and limited differentiation potential (Dontu et al., 2003; Ponti et al., 2005). In our hands, B40% of tumour specimens produced viable sphere cultures with a frequency of sphere forming cells ranging between Results 0.1% and 1% of the total cell suspension (data not shown). We examined the expression of Sox2 in Expression of pluripotency-associated genes in natural mammospheres from three tumour specimens using breast tumours real-time PCR (RT–PCR). As seen in Figure 1d, Sox2 Oct4, Sox2 and Nanog are considered the master expression was observed in two out of three sphere regulator genes of pluripotency and are switched off in cultures (specimens #207 and #208). We did not detect adult tissues, except for Sox2 that is expressed in neural Sox2 expression in specimen #220 in uncultured tumour progenitors (Zappone et al., 2000). However, reexpres- cells nor in tumour spheres. This result suggests that in

Oncogene Sox2 activation in breast cancer stem cells O Leis et al 1356

Figure 1 Expression of pluripotency-related genes in natural human breast tumours. (a) Upper set: representative staining of natural breast tumours for Oct4, Nanog, Sox2 and b-catenin by immunohistochemistry. Lower set: staining control for each marker. (b1) Quantitative scoring of marker-positive cells. Number of tumour patients for each marker is indicated. (b2) TNM stage classification of tumours and quantification of Sox2-positive cells. (c) Representative staining of two breast tumour specimens showing location of Sox2-positive cells (indicated by arrows) in the intra-ductal area. Quantification of the number of Sox2-positive cells is represented. Statistical significance was tested by Student t-test analysis (P ¼ 0.0347). (d) Sox2 expression is induced in tumour spheres (indicated by arrows) obtained from tumour cultures, assessed by RT–PCR from total RNA. Data are represented as DDCT value compared with a standard GAPDH expression control as described (Livak and Schmittgen, 2001). As a control, expression of Sox2 was assessed in total RNA obtained from a human embryonic stem cell culture (kindly provided by Dr R Sanchez Pernaute, Fundacio´n Inbiomed, Spain).

conditions where stem cell properties are revealed, such associated genes in breast carcinoma cell lines. Existence as the tumour sphere assay, Sox2 is expressed. of a discrete tumour-initiating population in breast As the availability and size of tumour samples is carcinoma cell lines has been already described (Patra- limited, we proceeded to examine pluripotency- wala et al., 2005; Engelmann et al., 2008; Fillmore and

Oncogene Sox2 activation in breast cancer stem cells O Leis et al 1357 Table 1 Comparison of Sox2 expression according to tumour subtype Kuperwasser, 2008). We therefore examined the ability of MCF7 breast carcinoma cells to grow as tumour Sox2 spheres, measuring the ability to form mammospheres in Negative Positive serial passages and to make clonal spheres from single cells, reflecting the stem cell-like properties of self- % (n) % (n) renewal and asymmetric division (Figure 2a). Indeed, Luminal A 86.7 (91) 13.3 (14) the mammosphere assay has been demonstrated in Luminal B 72.2 (13) 27.8 (5) MCF7 cells to enrich and propagate cells with enhanced Her2 69.2 (9) 30.8 (4) tumour-initiating ability (Ao et al., 2011). We used Basal 90.9 (20) 9.1 (2) a commercial antibody array that allows assessing Total 84.2 (133) 15.8 (25) the expression of Oct4, Sox2, Nanog, and other

Figure 2 Expression of pluripotency-related genes in mammospheres derived from MCF7 breast carcinoma cell line. (a1) Representative micrograph of a 7-day mammosphere culture from MCF7 breast carcinoma cell line. Quantification of the proliferation of sphere cultures upon serial passage is represented. Sphere passages are named T to distinguish from regular adherent culture passage (usually referred as P). (a2) Representative sequence of sphere formation from single cells at limiting dilution in 96-well plates. On average 30% of the cells from a typical MCF7 breast carcinoma cell culture were capable of producing clonal spheres. (b) Expression of pluripotency-related genes in total cell extracts using a Proteome Profiler Human Pluripotent Stem Cell Array (cat#ARY010, R&D Systems) in whole-cell extracts. A representative experiment is shown. Quantification of the arrays was done using Image-J software, and protein expression normalized to the average of the positive and negative controls provided in every array (dots A1–A2, A7–A8, F1–F2 for positive control and E7–E8 for negative control). Data are represented as fold change in expression calculated comparing the normalized expression value of mammospheres to parental unselected cell cultures (named ‘Stock’). (c) Immunofluorescence staining of Sox2-positive cells in representative mammospheres upon serial passages (red). Cell nucleus was stained using Hoescht 33342 stain (blue). Spheres are shown as  40 magnification. Achievement of internal sphere permeabilization is demonstrated as counterstaining with an antibody against E-Cadherin reveals cell-cell contacts in the inner part of the sphere (bottom- right panel). (d) Validation of the results obtained in the antibody array by immunoblotting against Sox2 in nuclear extracts obtained from parental adherent cultures (lane 1), T2 spheres (lane 2) and human embryonic stem cell culture as control (lane 3). Immunoblotting for Sam68 was performed as loading control.

Oncogene Sox2 activation in breast cancer stem cells O Leis et al 1358 pluripotency-associated genes in a single test (Figure 2b) protein controlled by an IRES element. Expression of to compare protein extracts obtained from serial Sox2 induced the generation of approximately three-fold passages of tumour spheres. Extracts from human more spheres than the empty vector control (Figure 3a). embryonic stem cells (kindly provided by R. Sanchez Interestingly, this effect was reversible and contingent Pernaute, Inbiomed, Spain) were used as control where to the activity of the plasmid, suggesting that the ectopic expression of Oct4, Nanog and Sox2 was readily expression of Sox2 was responsible for the increase in detected (data not shown). We did not observe expres- sphere number. To test this possibility, we generated sion of Oct4, Nanog or Sox2 in the stock control MCF7 cells with stable expression of Sox2 through population, however, induction of Sox2 protein but not lentiviral transfer (mCitrinie-P2A-Sox2 lentiviral vector) Oct4 or Nanog upon sphere formation was observed. and observed a marked increase in sphere formation Detection of Sox2 by immunoblotting in nuclear compared with mock-infected cells and those main- extracts from mammospheres confirmed the results tained in subsequent sphere passages (Figure 3b). obtained with the antibody array (Figure 2d). These We then used RNA interference silencing with double- results are consistent with the observations we obtained stranded RNA oligonucleotides (s13294 -si1-, s13295 using the fresh breast tumour specimens. -si2- and controls, Applied Biosystems, Carlsbad, CA, Next we stained mammospheres by immunofluores- USA) to knockdown Sox2 expression for testing whether cence in different passages to locate Sox2-positive cells Sox2 expression was necessary to induce tumour spheres. (Figure 2c). We did not observe Sox2 staining from Over 90% silencing was achieved with both siRNAs parental adherent cultures despite detecting positive used, as confirmed by immunoblotting (Figure 3c). expression in control extracts by immunoblotting, likely When tested for mammosphere formation, we observed due to insufficient sensitivity of the immunofluorescence an evident reduction of sphere formation compared with technique. We observed that only some of the cells in the control siRNA (50% reduction for -si1- and almost tumour spheres stained positive for Sox2 (2.9% for T1, compete blockage for -si2-) indicating that Sox2 5.7% for T2, 17.7% for T3, 12.1% for T4, 24.7% for T5 expression is necessary for mammosphere formation. and 32.0% for T6) and those cells located preferentially We then tested the tumour initiation ability of MCF7 to the periphery of the spheres. This effect was not due cells silenced for Sox2 expression in mouse xenograft to insufficient access of the antibody to the centre of the assays (Figure 3d and Supplementary Figure 2). As spheres, as the inner cells were readily detected by observed in Figure 3d, silencing of Sox2 expression with E-Cadherin staining (Figure 2c lower right panel). siRNA reduced the size of the tumours produced, as Therefore, it is tempting to speculate that those few compared with siRNA control cells, indicating that Sox2-positive cells represent the CSCs and/or their Sox2 facilitates tumour initiation. immediate progeny. Essentially, the same results were obtained using T47D cells, a different breast carcinoma cell line (Supplementary Figure 1). Sox2 promoter is induced upon sphere formation Regulation of Sox2 expression is poorly understood. An area located between positions À528 and þ 238 from Sox2 is necessary and sufficient for sphere formation the transcription start site is considered as the core In order to test whether Sox2 was conferring stem cell proximal promoter region (Miyagi et al., 2006). properties to the breast carcinoma cells, we used the Additionally, an upstream enhancer centred between pCAGSS-Sox2-GFP construct to transiently express À3444 and À3833 has an active role in controlling Sox2 in MCF7 cells. This vector allows the monitoring expression of Sox2 in the reprogramming of oligoden- of plasmid activity by expression of the green fluorescent drocyte precursors (Kondo and Raff, 2004) and in

Figure 3 Sox2 is sufficient and necessary for sphere formation. (a) Transient transfection of Sox2 with pCAGSS control or pCAGSS- Sox2 expression vectors. A schematic representation of the structure of the vector is shown. Twenty-four hours after transfection, cells were seeded in non-adherent conditions to allow for mammosphere formation and the number of spheres was counted after 7 days. Then spheres were disaggregated and reseeded to allow for T2 mammosphere formation. The number of T2 mammospheres was compared after 7 days. Expression of the plasmids was followed by green fluorescent protein expression measured in a flow cytometer (right panel) at every passage. (b) Stable expression of Sox2 was achieved through lentiviral transfer of Sox2-P2A-mCitrine lentiviral construct (described under Materials and methods). mCitrine-positive cells were sorted in a FACS Aria III cell sorter (Becton- Dickinson, Franklin Lakes, NJ, USA) to generate homogeneous cell cultures. Cells were seeded as described in (a) and the number of spheres was compared in each passage of Sox2 expressing cultures to that of mock-infected cultures. Right panel demonstrates Sox2 expression by immunoblotting using nuclear extracts from N-Tera2 cells as control and Sam68 expression as loading control. (c) siRNA-mediated silencing of Sox2 prevented mammosphere formation. Silencing was achieved using transient transfection of commercial siRNA oligonucleotides and appropriate controls. Forty-eight hours after transfection, cells were seeded in non-adherent conditions to allow for mammosphere formation and the number of spheres was compared after 7 days. Results are expressed as percentage of number of spheres compared with that of the control siRNA (black bars). Silencing of Sox2 was demonstrated by immunoblotting (right panel), as control Sam68 expression was not affected by siRNA treatment. A immunoblotting corresponding to a representative experiment of three is shown. Silencing quantification was achieved through scanning of the immunoblots and quantification of the digital images, results were expressed as percentage of Sox2 expression compared with the siRNA controls normalized by loading control (white bars). (d) Sox2 knockdown reduced tumour initiation in vivo. In all, 105 Sox2-silenced MCF7 cells (si2 sequence) or control siRNA cells were injected subcutaneously into both flanks of three mice per condition and tumour formation observed. Tumour size is expressed as tumour volume measured using a Vernier calliper as described in Materials and methods. Silencing of Sox2 was demonstrated by immunoblotting (not shown).

Oncogene Sox2 activation in breast cancer stem cells O Leis et al 1359

Oncogene Sox2 activation in breast cancer stem cells O Leis et al 1360

Figure 4 The distal enhancer of Sox2 promoter is induced upon sphere formation. (a) Schematic representation of Sox2 promoter structure indicating the proximal core promoter region and the location of the distal enhancer. Reporter constructs are specified. (b) In all, 5 mg of PGL3-Luc control, Sox2p-core-Luc and Sox2p-enhancer-Luc luciferase reporter vectors were transfected into MCF7 cells. Twenty-four hours after transfection, the culture was split in two, one part was seeded in 2D adherent culture and the other part was cultured in non-adherent culture conditions to allow mammosphere formation. After 48 h, cells were harvested and luciferase activity was compared, normalized by protein concentration in the extracts. Results are expressed as fold induction of sphere culture reporter activity above adherent culture control.

pluripotent stem cells (Tomioka et al., 2002; see changes, the stem cell self-renewing property. To Figure 4a for a schematic representation of the Sox2 investigate this issue we examined the presence of promoter). We transfected MCF7 and T47D cells with stem-like cells by assessing the expression of the major luciferase reporter vectors for the upstream distal stem cell pluripotency regulators, Oct4, Sox2 and Nanog enhancer and the core promoter to test whether (Jaenisch and Young, 2008), in human natural breast induction of Sox2 protein was achieved through tumours. We observed that Sox2 is expressed at the transcriptional activation of Sox2 promoter. As seen in protein level in breast tumours, as expected (Rodriguez- Figure 4b and Supplementary Figure 2, the Sox2 core Pinilla et al., 2007; Chen et al., 2008). However, Sox2 promoter region reporter was not significantly activated expression was more prevalent in the early stages of upon sphere formation whereas the distal enhancer tumour development and in the ductal areas of tumours region was strongly activated (seven-fold on average). that still showed intact ductal structures, as is the case of Taken together, these results suggest that reactivation of lung cancer (Sholl et al., 2010). During the preparation Sox2 expression in breast carcinoma cells upon sphere of this manuscript, Lengerke et al. (2011) reported the formation may be controlled at the promoter level, expression of Sox2 in early-stage breast carcinoma, similarly as it is activated in pluripotent stem cells. similar to the results we have obtained. Interestingly, they showed higher expression of Sox2 in ductal carcinoma as opposed to invasive carcinoma (44% versus 28%). Together, these results indicate that Sox2 Discussion may be expressed in the initial phases of tumourigenesis and then is lost as the tumour develops. The presence of CSCs in breast tumours is likely one of Spontaneous human breast tumours are predomi- the main reasons why current oncologic therapies are nantly differentiated luminal subtype (68%) while poorly effective in preventing tumour progression, undifferentiated tumours comprise only 24% of the metastasis and recurrence (Shafee et al., 2008; Tanei tumours (Callagy et al., 2003). It is common to assume et al., 2009; Cirenajwis et al., 2010), so elimination of that the tumour phenotype reflects the cell of origin of CSCs may become a necessary step for an effective cure. that tumour, thus, most natural occurring tumours The target of malignant transformation could be a tissue would be originated in the differentiated cells. Sox2 stem cell, a progenitor cell, or a terminally differentiated expression has been reported from normal mammary cell that acquires, through mutations and epigenetic stem cells when cultured as mammospheres and lost as

Oncogene Sox2 activation in breast cancer stem cells O Leis et al 1361 the cells differentiate in vitro (Simoes et al., 2010). This may act as a regulator of stem cell self-renewal in vivo assay is used as a stem-like functional assay that allows (Cicalese et al., 2009). Therefore, it is possible that the the propagation of mammary epithelial and breast acquisition of the self-renewal ability of CSCs arises by tumour cells in an undifferentiated state based on their some of the mechanisms shared with those involved in ability to proliferate in suspension (Dontu et al., 2003; pluripotency reprogramming, in cells that harbour Ponti et al., 2005; Farnie et al., 2007) and has been used permissive mutations, such as inactivation of the p53 for cancer stem-like specific drug screening in MCF7 pathway. This mechanism explains the generation of cells (Zhou et al., 2007; Ao et al., 2011). We did observe CSCs from committed progenitors with limited differ- induction of Sox2 expression in mammospheres ob- entiation potential responsible for the heterogeneous tained from human breast tumour cell cultures, and subtypes observed in natural tumours. We did not from MCF7 and T47D cells. Moreover, we did not observe induction of Oct4 or Nanog, therefore, a full observe expression of any of the other pluripotency- pluripotency programme cannot be established. The associated genes in MCF7-derived spheres, consistently ability of non-stem cells to convert to CSCs implies with the results obtained in human tumours. Despite intrinsic plasticity at the level of the cell of origin of the some reports claiming expression of Oct4 and Nanog in tumours. In this regard, transformation of an immorta- breast carcinoma cell lines, expression of several lized breast epithelial cell line (MCF10A) by the pseudogenes have been reported for both Oct4 and oncogene Src generated cells with features of CSCs only Nanog (Suo et al., 2005; Zhang et al., 2006; Cantz et al., when interleukin-6 was added to the medium, suggesting 2007; Ambady et al., 2010), therefore, particular care that CSCs can be originated from non-CSC transformed must be taken to avoid PCR amplification of those cells (Iliopoulos et al., 2011). Similar results were pseudogenes. We base our conclusions in antibody- obtained attending to in vitro culture of floating human based assays, thus, looking specifically at the protein mammary epithelial cells, where cells with stem-like level to avoid pseudogene detection. As opposed to features arise from FACS (fluorescence-activated cell spheres derived from glioma cell lines (Ghods et al., sorting)-sorted differentiated epithelial cells (Chaffer 2007), only a few cells in the mammospheres were stained et al., 2011). We propose that spontaneous dediffer- positive for Sox2, thus it is tempting to speculate that the entiation in vivo may involve reactivation of one or more tumour-initiating property resides in those cells. Further pluripotency-associated factors, such as Sox2 expression experiments will be needed to specifically isolate Sox2- in breast cancer, consistent with our observations. What positive cells and test their tumour generating ability. the mechanisms for reactivation are, how are they In order to verify the contribution of Sox2 to different from a full pluripotency programme and how mammosphere formation, we over-expressed Sox2 and contextual signals from the microenvironment (such as observed that Sox2 expression increased the ability to induction of EMT (Mani et al., 2008) participate in this generate spheres. Interestingly, this effect was dependent process, will shed light into how tumours are initiated on the continuous expression of Sox2 indicating that Sox2 and how to control this process. expression was sufficient to induce this stem-like feature. How Sox2 expression is reactivated remains unclear. Knockdown of Sox2 expression prevented the formation Sox2 promoter consists of two main regulatory regions, of spheres; thus, Sox2 expression was not only sufficient the core promoter located at the transcription initiation but also necessary to form spheres. Furthermore, when site and an upstream distal enhancer. Our results Sox2-silenced MCF7 cells were tested for tumour indicate that it is precisely this upstream distal enhancer initiation in mouse xenograft models, we observed a that is activated in mammospheres, as a luciferase marked reduction in the size of the tumours formed, reporter with this enhancer is strongly induced upon suggesting that Sox2 participates in tumour initiation sphere formation. The upstream enhancer was strongly in vivo. This result is consistent with data reported on activated in undifferentiated embryonic stem cells and glioblastoma where Sox2 knockdown prevents glioma its activity was lost upon differentiation (Tomioka et al., stem cells to initiate tumours (Gangemi et al.,2009). 2002), it is the same behaviour as observed in multi- The process of in vitro pluripotency induction presents potent neural stem cells (Miyagi et al., 2004; Miyagi several parallelisms with tumour generation. Five et al., 2006). In embryonic stem cells, a non-consensus groups independently (Hong et al., 2009; Kawamura octamer-binding site in the enhancer is occupied by et al., 2009; Li et al., 2009; Marion et al., 2009; Utikal Oct3/4-Sox2 dimers, however, as we did not observe et al., 2009) observed that p53 poses a barrier for Oct4 expression in mammospheres, a different mechan- pluripotency induction. Mutations that inactivate the ism must operate. Recently, a possible implication of the p53 tumour suppressor network occur in most natural transcription factor NF (nuclear factor)-kB has been human tumours, thus, remarkable similarities must exist proposed, although no direct regulation was reported between the genetic and epigenetic processes involved in (Liu et al., 2010). Kondo and Raff (Kondo and Raff, the generation of cancer, and the acquisition of a less 2004) demonstrated that Sox2 reactivation in the differentiated phenotype and self-renewal ability in conversion of oligodendrocyte precursors depends on induced pluripotency. Loss of p53 confers stem-like the recruitment of the protein Brca1 to this enhancer, properties in committed progenitors in the haemato- together with the chromatin remodelling protein Brm. poietic system (Akala et al., 2008), and loss of p53 These complexes have a critical role in the maintenance increases the size of the mammary repopulating of the pluripotent state by regulating occupancy of compartment in mice by 10-fold, suggesting that p53 pluripotency-related gene promoters (Kidder et al.,

Oncogene Sox2 activation in breast cancer stem cells O Leis et al 1362 2009). As Brca1 is a well-known tumour suppressor gene Council, London, UK). For immunoblotting, immunofluores- that when mutated predisposes women to breast and cence and immunohistochemistry anti-human Sox2 polyclonal ovarian cancer (Miki et al., 1994), a possibility is that antibody (Neuromics GT15098), anti-human Oct4 polyclonal loss of Brca1 mediates tumour initiation by losing the antibody (Santa Cruz sc-8629, Santa Cruz, CA, USA), anti- control on Sox2 expression, establishing a direct link human Nanog polyclonal antibody (R&D Systems AF1997), between Sox2 expression and breast cancer initiation. anti-human b-catenin monoclonal antibody (Millipore 05665, Billerica, MA, USA), anti-human E-cadherin polyclonal anti- Nevertheless, this model needs further testing. body (Cell Signaling FAB18381A, Danvers, MA, USA) and anti-human Sam68 polyclonal antibody (Santa Cruz sc-333) were used as indicated. Materials and methods Immunohistochemistry Tumour sample collection Immunohistochemical staining on tumour sections was per- A total of 158 cases of breast tumour samples were collected at formed using Dako Autostainer automatic universal slide the Pathology Department of Onkologikoa (San Sebastian, stainer. Antibodies, dilutions, antigen-retrieval methods and Spain) from patients that underwent curative surgical resection scoring systems for ER, PR, HER2, CK5/6, EGFR and ki67 (for tumour data see Supplementary File 1). Institutional are part of the routine histopathological determination at the review board approval and expressed informed consent were Pathology Department of Onkologikoa (San Sebastian, obtained from all patients before sample collection. Formalin- Spain). Cases were considered positive for Sox2 when any fixed paraffin-embedded sections of tumours were stained unequivocal neoplastic cell displayed definite nuclear staining. routinely with haematoxylin-eosin and reviewed by two Positive (as indicated) and negative controls were included in senior pathologists (RR and KE) to determine the histological each slide run. type according to WHO breast carcinoma histological classification criteria (2003), and clinical stage according to the UICC TNM classification (2003). Tumours were sub- Real-time PCR divided as: basal-like (ER, PR and HER2-negative and CK5/6 Total RNA was obtained using RNAqueous-Micro kit (Am- and/or EGFR-positive), luminal A (ER and/or PR-positive bion, Austin, TX, USA) following manufacturer’s instructions. and HER2-negative), luminal B (ER and/or PR-positive and Power SYBER Green PCR Master Mix (Applied Biosystems) HER2-positive) and HER2-positive subgroups. The data and was used to amplify the corresponding genes with primers for tumour samples used were provided by the Basque Biobank human Sox2 (forward: 50-GGGAAATGGGAGGGGTGCAA for Research (http://www.basquebiobank.org), supported by AAGAGG-30,reverse:50-TTGCGTGAGTGTGGATGGGGA Basque Foundation for Health Innovation and Research. TTGGTG-30). Amplification of human GAPDH (Hs_GAP DH_2_SG QuantiTect Primer Assay QT01192646, Qiagen, Cell culture Hilden, Germany) was used as loading control. RT–PCR was Fresh tumour samples were cut up into small pieces and minced performed in a StepOnePlus Real-Time PCR System. completely using scissors. Tumour pieces were enzymatically digested with Collagenase IA (Sigma, St Louis, MO, USA) at Immunofluorescence the concentration of 2 mg/ml and incubated at 37 1Cfor1h. Mammospheres were collected by centrifugation in a Then, the digested sample was filtered through a 40-mm Cytospin (Thermo, Waltham, MA, USA) onto glass slides. cell strainer and washed with Hank’s balanced salt solution. Slides were fixed in 4% paraformaldehyde for 1 h at room For sphere formation, cells were cultured at an initial density temperature in wet chamber, washed twice in phosphate- of 50 000 cells/ml in serum-free Dulbecco’s modied Eagle’s buffered saline (PBS) and allowed to permeabilize for 1 h at medium supplemented with 1% L-glutamine, 1% penicillin/ room temperature with PBS þ 0.3% Triton X-100 and streptomycin, 30% F12 (Sigma), 2% B27 (Invitrogen, Carlsbad, blocked. Slides were then incubated with primary antibodies CA, USA), 20 ng/ml EGF (Sigma) and 20 ng/ml FGFb overnight and washed thrice in PBS, and appropriate (Invitrogen) in non-adherent culture plates. For subsequent fluorofore labelled secondary antibodies were added at a sphere formation, cells were seeded at a density of 5 000 cells/ml. dilution of 1/250. Hoescht 33342 dye (Sigma, St Louis, MO, MCF-7 and T47D breast carcinoma cell lines were obtained USA) was used to reveal nuclear DNA. Immunofluorescence directly from the ATCC (Manassas, VA, USA) and grown in was visualized in a Zeiss LSM510 confocal laser-scanning Dulbecco’s modied Eagle’s medium (Gibco, Carlsbad, CA, microscope (Zeiss, Jena, Germany). USA) supplemented with 10% fetal bovine serum (Sigma) at 37 1Cina5%CO2 incubator. For mammosphere formation, single-cell suspensions were plated in six-well tissue culture Immunoblotting plates covered with poly-2-hydroxyethyl-methacrylate (Sigma) Nuclear extracts (prepared as in Martin et al. (2001)) were to prevent cell attachment at a density of 1000 cells/ml in blotted onto a polyvinylidene fluoride membrane (Millipore) the medium described above. The medium was made semi-solid and blocked in Tris-buffered saline with 0.2% Tween 20 and by the addition of 0.5% Methylcellulose (R&D Systems, 5% milk powder. Immunoblotting for Sox2 and Sam68 was Minneapolis, MN, USA) to prevent cell aggregation. Mammo- carried out using primary antibodies at a 1:2000 dilution and spheres were collected by gentle centrifugation after 7 days, and appropriate secondary antibody coupled to horseradish dissociated enzymatically (5 min in 1:1 trypsin/Dulbecco’s peroxidase (1:10 000 dilution). SuperSignal West Pico Chemi- modied Eagle’s medium solution at 37 1C) and mechanically luminescent Substrate developing kit (Thermo) was used to by passing through a 25G needle. Single cells were replated at a reveal the relevant bands. density of 1000 cells/ml for subsequent passages. Transfection, gene silencing and lentiviral gene transfer Plasmids and antibodies For transient transfection, MCF-7 cells were cultured in six- pCAGSS-Sox2 and empty pCAGSS control expression vectors well tissue culture plates and transfected when at 70% were kindly provided by A Matheu (Medical Research confluence using Lipofectamine Plus or Oligofectamine

Oncogene Sox2 activation in breast cancer stem cells O Leis et al 1363 (Invitrogen) according to manufacturer’s instructions. Gene Luciferase reporter assays silencing experiments were performed using specific commer- MCF-7 cells were transfected with 5 mg of pGL3 Luc control cial siRNA sequences for human Sox2 (s13294 and s13295, (Promega, Madison, WI, USA), pGL3-Sox2p-enhancer-Luc Applied Biosystems), and appropriate siRNA controls. For (kindly provided by T Kondo, RIKEN, Japan) or pGL3- stable expression of Sox2, lentiviral transfer was used using Sox2p-core-Luc (kindly provided by JM Sanchez-Puelles, CIB, the mCitrinie-P2A-Sox2 lentiviral vector (kindly provided by Madrid, Spain) reporter plasmids using Lipofectamine Plus E Papapetrou, Memorial Sloan-Kettering Cancer Center, NY, (Invitrogen). Luciferase activity was measured in duplicate USA; Papapetrou et al., 2009). with the Glomax 20/20 luminometer (Promega) and the results were expressed as fold induction above control. Mouse xenograft assays Female 6-week-old athymic nude mice (Balb/c Nu/Nu) were purchased from Charles River, and were housed in specifically designed pathogen-free isolation animal facility. All animal Conflict of interest procedures were performed in accordance with institutional animal care and use guidelines. For injection, MCF7 cells of The authors declare no conflict of interest. each condition were resuspended in 10 ml of the same media at a concentration of 1 Â 106 cells/ml. Mice were subcutaneously inoculated with the MCF7 cell line in 200 ml of culture medium with matrigel in both mammary fat pads. In all, 105 MCF7 Acknowledgements cells with Sox2 knockdown were inoculated in the right mammary fat pad, with their respective control in the left The Regulation of Cell Growth Laboratory is supported by mammary fat pad. Mice were weighed and the inoculation sites grants from Obra Social Kutxa, Fundacio´nMe´dica Mutua were inspected by palpation at weekly intervals. When Madrilen˜a, Gobierno Vasco (Saiotek program and Consejerı´a tumours become detectable manually, the growth rates were de Educacio´n PI2010-25) and Instituto de Salud Carlos III determined by weekly measurement of two diameters of the Accio´n Estrate´gica en Salud (PI2010-01035). We thank tumour with a Vernier calliper. The tumour volume was Izaskun Beloqui, Andres Pavon and Maria Diaz for technical estimated as the volume of an ellipse using the following support. We thank the Flow Cytometry Unit at Inbiomed for formula: V ¼ 4/3 Â (a/2) Â (b/2)2, where ‘a’ and ‘b’ correspond extensive aid in flow cytometric analysis of fresh tumour cells. to the longest and shortest diameter, respectively. Animals We thank our colleagues at Inbiomed for helpful discussion. were euthanized when their tumours were harvested, fixed and We thank Dr R Sanchez-Pernaute and Dr L Vellon for critical embedded in paraffin. reading of the manuscript.

References

Abraham BK, Fritz P, McClellan M, Hauptvogel P, Athelogou M, of self-renewing divisions in mammary stem cells. Cell 138: Brauch H. (2005). Prevalence of CD44+/CD24À/low cells in breast 1083–1095. cancer may not be associated with clinical outcome but may favor Cirenajwis H, Smiljanic S, Honeth G, Hegardt C, Marton LJ, distant metastasis. Clin Cancer Res 11: 1154–1159. Oredsson SM. (2010). Reduction of the putative CD44+CD24À Akala OO, Park IK, Qian D, Pihalja M, Becker MW, Clarke MF. breast cancer stem cell population by targeting the (2008). Long-term haematopoietic reconstitution by Trp53-/- polyamine metabolic pathway with PG11047. Anticancer Drugs p16Ink4a-/-p19Arf-/- multipotent progenitors. Nature 453: 228–232. 21: 897–906. Al-Hajj M, Wicha MS, Benito-Hernandez A, Morrison SJ, Clarke Comtesse N, Zippel A, Walle S, Monz D, Backes C, Fischer U et al. MF. (2003). Prospective identification of tumorigenic breast cancer (2005). Complex humoral immune response against a benign tumor: cells. Proc Natl Acad Sci USA 100: 3983–3988. frequent antibody response against specific antigens as diagnostic Ambady S, Malcuit C, Kashpur O, Kole D, Holmes WF, Hedblom E targets. Proc Natl Acad Sci USA 102: 9601–9606. et al. (2010). Expression of NANOG and NANOGP8 in a variety of Dontu G, Abdallah WM, Foley JM, Jackson KW, Clarke MF, undifferentiated and differentiated human cells. Int J Dev Biol 54: Kawamura MJ et al. (2003). in vitro propagation and transcrip- 1743–1754. tional profiling of human mammary stem/progenitor cells. Genes Ao A, Morrison BJ, Wang H, Lopez JA, Reynolds BA, Lu J. (2011). Dev 17: 1253–1270. Response of estrogen receptor-positive breast cancer tumorspheres Engelmann K, Shen H, Finn OJ. (2008). MCF7 side population cells to antiestrogen treatments. PLoS ONE 6: e18810. with characteristics of cancer stem/progenitor cells express the Callagy G, Cattaneo E, Daigo Y, Happerfield L, Bobrow LG, Pharoah tumor antigen MUC1. Cancer Res 68: 2419–2426. PD et al. (2003). Molecular classification of breast carcinomas using Farnie G, Clarke RB, Spence K, Pinnock N, Brennan K, Anderson tissue microarrays. Diagn Mol Pathol 12: 27–34. NG et al. (2007). Novel cell culture technique for primary ductal Cantz T, Key G, Bleidissel M, Gentile L, Han DW, Brenne A et al. carcinoma in situ: role of Notch and epidermal growth factor (2007). Absence of OCT4 expression in somatic tumor cell lines. receptor signaling pathways. J Natl Cancer Inst 99: 616–627. Stem Cells 26: 692–697. Fillmore CM, Kuperwasser C. (2008). Human breast cancer cell Chaffer CL, Brueckmann I, Scheel C, Kaestli AJ, Wiggins PA, lines contain stem-like cells that self-renew, give rise to phenotypi- Rodrigues LO et al. (2011). Normal and neoplastic nonstem cells cally diverse progeny and survive chemotherapy. Breast Cancer Res can spontaneously convert to a stem-like state. Proc Natl Acad Sci 10: R25. USA 108: 7950–7955. Gangemi RM, Griffero F, Marubbi D, Perera M, Capra MC, Chen Y, Shi L, Zhang L, Li R, Liang J, Yu W et al. (2008). The Malatesta P et al. (2009). SOX2 silencing in glioblastoma tumor- molecular mechanism governing the oncogenic potential of SOX2 in initiating cells causes stop of proliferation and loss of tumorigeni- breast cancer. J Biol Chem 283: 17969–17978. city. Stem Cells 27: 40–48. Cicalese A, Bonizzi G, Pasi CE, Faretta M, Ronzoni S, Ghods AJ, Irvin D, Liu G, Yuan X, Abdulkadir IR, Tunici P et al. Giulini B et al. (2009). The tumor suppressor p53 regulates polarity (2007). Spheres isolated from 9L gliosarcoma rat cell line possess

Oncogene Sox2 activation in breast cancer stem cells O Leis et al 1364 chemoresistant and aggressive cancer stem-like cells. Stem Cells 25: Miyagi S, Nishimoto M, Saito T, Ninomiya M, Sawamoto K, Okano 1645–1653. H et al. (2006). The Sox2 regulatory region 2 functions as a neural Ginestier C, Hur MH, Charafe-Jauffret E, Monville F, Dutcher J, stem cell-specific enhancer in the telencephalon. J Biol Chem 281: Brown M et al. (2007). ALDH1 is a marker of normal and 13374–13381. malignant human mammary stem cells and a predictor of poor Miyagi S, Saito T, Mizutani K, Masuyama N, Gotoh Y, Iwama A clinical outcome. Cell Stem Cell 1: 555–567. et al. (2004). The Sox-2 regulatory regions display their activities Gure AO, Stockert E, Scanlan MJ, Keresztes RS, Jager D, Altorki NK in two distinct types of multipotent stem cells. Mol Cell Biol 24: et al. (2000). Serological identification of embryonic neural proteins 4207–4220. as highly immunogenic tumor antigens in small cell lung cancer. Nakagawa M, Koyanagi M, Tanabe K, Takahashi K, Ichisaka T, Proc Natl Acad Sci USA 97: 4198–4203. Aoi T et al. (2008). Generation of induced pluripotent stem cells Hong H, Takahashi K, Ichisaka T, Aoi T, Kanagawa O, Nakagawa M without Myc from mouse and human fibroblasts. Nat Biotechnol 26: et al. (2009). Suppression of induced pluripotent stem cell 101–106. generation by the p53-p21 pathway. Nature 460: 1132–1135. Okita K, Nakagawa M, Hyenjong H, Ichisaka T, Yamanaka S. (2008). Iliopoulos D, Hirsch HA, Wang G, Struhl K. (2011). Inducible Generation of mouse induced pluripotent stem cells without viral formation of breast cancer stem cells and their dynamic equilibrium vectors. Science 322: 949–953. with non-stem cancer cells via IL6 secretion. Proc Natl Acad Sci Papapetrou EP, Tomishima MJ, Chambers SM, Mica Y, Reed E, USA 108: 1397–1402. Menon J et al. (2009). Stoichiometric and temporal requirements of Jaenisch R, Young R. (2008). Stem cells, the molecular circuitry of Oct4, Sox2, Klf4, and c-Myc expression for efficient human pluripotency and nuclear reprogramming. Cell 132: 567–582. iPSC induction and differentiation. Proc Natl Acad Sci USA 106: Katoh M. (2010). Network of WNT and other regulatory signaling 12759–12764. cascades in pluripotent stem cells and cancer stem cells. Curr Pharm Patrawala L, Calhoun T, Schneider-Broussard R, Zhou J, Claypool K, Biotechnol 12: 160–170. Tang DG. (2005). Side population is enriched in tumorigenic, stem- Kawamura T, Suzuki J, Wang YV, Menendez S, Morera LB, Raya A like cancer cells, whereas ABCG2+ and ABCG2À cancer cells are et al. (2009). Linking the p53 tumour suppressor pathway to somatic similarly tumorigenic. Cancer Res 65: 6207–6219. cell reprogramming. Nature 460: 1140–1144. Ponti D, Costa A, Zaffaroni N, Pratesi G, Petrangolini G, Coradini D Kidder BL, Palmer S, Knott JG. (2009). SWI/SNF-Brg1 regulates self- et al. (2005). Isolation and in vitro propagation of tumorigenic renewal and occupies core pluripotency-related genes in embryonic breast cancer cells with stem/progenitor cell properties. Cancer Res stem cells. Stem Cells 27: 317–328. 65: 5506–5511. Kim JB, Greber B, Arauzo-Bravo MJ, Meyer J, Park KI, Zaehres H Rodriguez-Pinilla SM, Sarrio D, Moreno-Bueno G, Rodriguez-Gil Y, et al. (2009). Direct reprogramming of human neural stem cells by Martinez MA, Hernandez L et al. (2007). Sox2: a possible driver of OCT4. Nature 461: 649–643. the basal-like phenotype in sporadic breast cancer. Mod Pathol 20: Kondo T, Raff M. (2004). Chromatin remodeling and histone 474–481. modification in the conversion of oligodendrocyte precursors to Sanada Y, Yoshida K, Ohara M, Oeda M, Konishi K, Tsutani Y. neural stem cells. Genes Dev 18: 2963–2972. (2006). Histopathologic evaluation of stepwise progression of Lengerke C, Fehm T, Kurth R, Neubauer H, Scheble V, Muller F et al. pancreatic carcinoma with immunohistochemical analysis of gastric (2011). Expression of the embryonic stem cell marker SOX2 in epithelial transcription factor SOX2: comparison of expression early-stage breast carcinoma. BMC Cancer 11: 42. patterns between invasive components and cancerous or nonneo- Li H, Collado M, Villasante A, Strati K, Ortega S, Canamero M et al. plastic intraductal components. Pancreas 32: 164–170. (2009). The Ink4/Arf locus is a barrier for iPS cell reprogramming. Sattler HP, Lensch R, Rohde V, Zimmer E, Meese E, Bonkhoff H Nature 460: 1136–1139. et al. (2000). Novel amplification unit at chromosome 3q25-q27 in Li X, Lewis MT, Huang J, Gutierrez C, Osborne CK, Wu MF et al. human prostate cancer. Prostate 45: 207–215. (2008). Intrinsic resistance of tumorigenic breast cancer cells to Shafee N, Smith CR, Wei S, Kim Y, Mills GB, Hortobagyi GN et al. chemotherapy. J Natl Cancer Inst 100: 672–679. (2008). Cancer stem cells contribute to cisplatin resistance in Li XL, Eishi Y, Bai YQ, Sakai H, Akiyama Y, Tani M et al. (2004). Brca1/p53-mediated mouse mammary tumors. Cancer Res 68: Expression of the SRY-related HMG box protein SOX2 in human 3243–3250. gastric carcinoma. Int J Oncol 24: 257–263. Sholl LM, Barletta JA, Yeap BY, Chirieac LR, Hornick JL. (2010). Liu M, Sakamaki T, Casimiro MC, Willmarth NE, Quong AA, Sox2 protein expression is an independent poor prognostic Ju X et al. (2010). The canonical NF-\{kappa\}B pathway indicator in stage I lung adenocarcinoma. Am J Surg Pathol 34: governs mammary tumorigenesis in transgen. Cancer Res 70: 1193–1198. 10464–10473. Simoes BM, Piva M, Iriondo O, Comaills V, Lopez-Ruiz JA, Zabalza I Livak KJ, Schmittgen TD. (2001). Analysis of relative gene expression et al. (2010). Effects of estrogen on the proportion of stem cells in data using real-time quantitative PCR and the 2(-Delta Delta C(T)) the breast. Breast Cancer Res Treat 129: 23–35. method. Methods 25: 402–408. Suo G, Han J, Wang X, Zhang J, Zhao Y, Zhao Y et al. (2005). Oct4 Mani SA, Guo W, Liao MJ, Eaton EN, Ayyanan A, Zhou AY et al. pseudogenes are transcribed in cancers. Biochem Biophys Res (2008). The epithelial-mesenchymal transition generates cells with Commun 337: 1047–1051. properties of stem cells. Cell 133: 704–715. Takahashi K, Tanabe K, Ohnuki M, Narita M, Ichisaka T, Tomoda K Marion RM, Strati K, Li H, Murga M, Blanco R, Ortega S et al. (2007). Induction of pluripotent stem cells from adult human et al. (2009). A p53-mediated DNA damage response limits fibroblasts by defined factors. Cell 131: 861–872. reprogramming to ensure iPS cell genomic integrity. Nature 460: Tanei T, Morimoto K, Shimazu K, Kim SJ, Tanji Y, Taguchi T et al. 1149–1153. (2009). Association of breast cancer stem cells identified by aldehyde Martin AG, San-Antonio B, Fresno M. (2001). Regulation of nuclear dehydrogenase 1 expression with resistance to sequential paclitaxel factor kappa B transactivation. Implication of phosphatidylinositol and epirubicin-based chemotherapy for breast cancers. Clin Cancer 3-kinase and protein kinase C zeta in c-Rel activation by tumor Res 15: 4234–4241. necrosis factor alpha. J Biol Chem 276: 15840–15849. Tomioka M, Nishimoto M, Miyagi S, Katayanagi T, Fukui N, Niwa Meissner A, Wernig M, Jaenisch R. (2007). Direct reprogramming of H et al. (2002). Identification of Sox-2 regulatory region which is genetically unmodified fibroblasts into pluripotent stem cells. Nat under the control of Oct-3/4-Sox-2 complex. Nucleic Acids Res 30: Biotechnol 25: 1177–1181. 3202–3213. Miki Y, Swensen J, Shattuck-Eidens D, Futreal PA, Harshman K, Utikal J, Polo JM, Stadtfeld M, Maherali N, Kulalert W, Walsh RM Tavtigian S et al. (1994). A strong candidate for the breast and et al. (2009). Immortalization eliminates a roadblock during cellular ovarian cancer susceptibility gene BRCA1. Science 266: 66–71. reprogramming into iPS cells. Nature 460: 1145–1148.

Oncogene Sox2 activation in breast cancer stem cells O Leis et al 1365 Yamanaka S. (2008). Induction of pluripotent stem cells from Zeng YA, Nusse R. (2010). Wnt proteins are self-renewal factors for mouse fibroblasts by four transcription factors. Cell Prolif mammary stem cells and promote their long-term expansion in 41(Suppl 1): 51–56. culture. Cell Stem Cell 6: 568–577. Zappone MV, Galli R, Catena R, Meani N, De Biasi S, Mattei E et al. ZhangJ,WangX,LiM,HanJ,ChenB,WangBet al. (2006). NANOGP8 (2000). Sox2 regulatory sequences direct expression of a (beta)-geo is a retrogene expressed in cancers. FEBS J 273: 1723–1730. transgene to telencephalic neural stem cells and precursors of the Zhou J, Zhang H, Gu P, Bai J, Margolick JB, Zhang Y. (2007). NF- mouse embryo, revealing regionalization of gene expression in CNS kappaB pathway inhibitors preferentially inhibit breast cancer stem- stem cells. Development 127: 2367–2382. like cells. Breast Cancer Res Treat 111: 419–427.

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

Oncogene