Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

The Controversial Clinicobiological Role of Breast Cancer Stem Cells

The Controversial Clinicobiological Role of Breast Cancer Stem Cells Hindawi Publishing Corporation Journal of Oncology Volume 2008, Article ID 492643, 12 pages doi:10.1155/2008/492643 Review Article The Controversial Clinicobiological Role of Breast Cancer Stem Cells 1 2 1 Claudia Casarsa, Saro Oriana, and Danila Coradini Experimental Oncology Laboratory, Senology Center, Ambrosiana Clinic, Cesano Boscone, 20090 Milano, Italy Surgery Department, Senology Center, Ambrosiana Clinic, Cesano Boscone, 20090 Milano, Italy Correspondence should be addressed to Danila Coradini, danila.coradini@yahoo.it Received 27 July 2008; Revised 5 December 2008; Accepted 23 December 2008 Recommended by Meenhard Herlyn Breast cancer remains a leading cause of morbidity and mortality in women mainly because of the propensity of primary breast tumors to metastasize. Growing experimental evidence suggests that cancer stem cells (CSCs) may contribute to tumor progression and metastasis spread. However, despite the tremendous clinical potential of such cells and their possible therapeutic management, the real nature of CSCs remains to be elucidated. Starting from what is currently known about normal mammary stem/progenitor cells, to better define the cell that originates a tumor or is responsible for metastatic spread, this review will discuss experimental evidence of breast cancer stem cells and speculate about the clinical importance and implications of their evaluation. Copyright © 2008 Claudia Casarsa et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. 1. Introduction of the tumor cells, invasive breast cancers are categorized by the World Health Organization into 18 different histological Despite significant advances in diagnosis and clinical man- subtypes, each of them associated with a diverse clinical behavior. In addition to morphology, invasive breast cancers agement, breast cancer remains a leading cause of morbidity and mortality in women [1], mainly owing to the propensity can also be classified according to their proliferative potential of primary breast tumors to metastasize to regional and (evaluated, e.g., by Ki67 expression) or the presence of such distant sites such as lymph nodes, lung, liver, bone, and brain biological factors as hormone receptors (estrogen [ER] and [2]. When the primary tumor is detected and removed before progesterone [PgR]) or HER2/neu overexpression that are metastasis occurs, prognosis could be good and the chance currently used in clinical practice to predict the prognosis of disease-free survival is high. However, if cancer cells have and the response/resistance to cytotoxic and/or hormonal already begun to disseminate from the primary tumor and therapy [4–6]. Understanding the molecular causes of such a hetero- spread to other organs, current therapeutic strategies largely depend on the use of systemic cytotoxic drugs that frequently geneity is therefore of paramount importance not only result in severe side effects on the patient and, in many cases, for the development of new therapeutic approaches, but do not yield long-term success. also for a better knowledge of the biological bases of This clinical scenario is further complicated by the fact breast tumorigenesis and metastatic spread. To address these that invasive breast cancers exhibit a wide range of mor- questions, over the past decade, scientists have used inno- phological types, molecular profiles, and clinical behaviors. vative technical strategies and approached new intriguing Not only there is a large variation in the nature of cell types directions aimed to define the genetic and epigenetic profile between cancers, but even within a single tumor a significant of the single tumor and to better define the cell that originates heterogeneity in phenotype and genotype can be observed a tumor or is responsible for metastatic spread. In particular, [3]. Based on growth patterns and cytological characteristics investigators focused their attention on the hypothesis that 2 Journal of Oncology breast cancer may be a stem cell disease, arising from In fact, contrary to the dogma that epithelial cell survival tissue stem/progenitor cells or driven by cells with stem-cell is anchorage-dependent, single cell suspensions of human properties [7]. mammary epithelial cells, obtained by mechanic/enzymatic dissociation, surprisingly survived in suspension and gener- ated floating spherical colonies, termed nonadherent mam- 2. Isolation and Characterization of Mammary mospheres [15]. The mammospheres contained numerous Stem/Progenitor Cells undifferentiated cells that, once isolated from the cluster, were able to generate new multilineage colonies, when In the adult, cell loss, associated with the physiological tissue cultured under differentiating conditions, and, in 3D culture, turnover, is compensated by the activity of specific cells, to reconstitute a functional mammary gland. However, the termed stem cells, which are so defined by their ability intrinsic dynamics of such cytospheres, as conventionally to self-renew and to generate the entire repertoire of the assayed, introduces several confounders, as reported in a differentiated cells composing a given tissue. Since 1983, recent paper by Singec et al. [16] who underlined the need the existence of such stem cells has also been postulated to use more accurate conditions for assessing the clonality, in the mammary gland to explain the cellular dynamics number, and fate of stem cells. Although sphere formation underlying morphological changes throughout a woman’s may represent a useful culturing tool, it is not specific to life, particularly during and after pregnancy [8]. stem cell characterization; any dividing cell from virtually In general, the identification and purification of normal any tissue will form floating cell clusters, when cultured in a stem cells are difficult tasks because of the paucity of stem serum-free medium and on a nonadherent substrate, owing cells in the tissue of origin and to the lack of stemness- to a predominant intercellular adhesiveness. Spontaneous specific morphologic traits. Hence, animal models have been sphere fusion may occur in normal as well as in neoplastic used, and the murine model, in particular, has supplied sphere cultures. Furthermore, in agreement with Singec et al. relevant data to improve our understanding of the cell [16], we have observed that the mammospheres, supposedly biology of the mammary gland and to clarify the presence rich in multilineage progenitors, have a very short life span of stem/progenitor cell and differentiated cell compartments (about 3-4 weeks), making it difficult to define the cells in the mammary gland [9, 10]. However, the information composing them as real mammary stem cells, which are long- obtained from mice cannot be directly applied to humans lived by definition. On account of all these criticisms, this (as in the case of murine exclusive cell surface markers) or experimental approach, based on the ability of the supposed may not be promptly translated to humans, as known in mammary stem cells to generate clonal mammospheres the hematopoietic field, where an opposite surface-antigen under anchorage-independent culture conditions, has been + − profile was observed in humans (CD34 CD38 )withrespect − + defined “a surrogate stem cell assay” [17]. to mice (CD34 CD38 ). Several in vitro strategies have thus been developed to isolate and characterize human mammary 2.3. Fluorescence Methods. To overcome the challenge repre- stem/progenitor cells, based on the differential expression of sented by the limited availability of stemness markers and to some cell surface markers, the formation of mammospheres, take advantage of the ability of stem cells to extrude dyes, for and the use of fluorescent dyes. example, Hoechst-33342 DNA-binding dye or rhodamine because the overexpression of some membrane transporter 2.1. Cell Surface Markers. The first experimental evidence proteins, such as P-glycoproteins or breast cancer resistance of the existence of human mammary stem/progenitor cells proteins (BCRPs) [18, 19], fluorescent dyes have been used was obtained by the in vitro isolation of multipotent to identify and isolate by flow cytometry a small fraction of epithelial cells, from normal human adult breast, accord- cells supposed to be stem progenitors [20]. This cell fraction, ing to their different expression of MUC-1 glycoprotein, which amounts to around 0.2% of the total population, has CALLA/CD10, and epithelial-specific antigen (ESA). Using been called side population (SP). these markers, two epithelial-cell progenitor populations However, several criticisms to such a sorting technique were distinguished, corresponding to the two components have been raised in the recent years principally concerning of normal mammary gland (myoepithelial cells, forming the the toxicity of dyes used in the analysis [21] and the basal layer of ducts, and epithelial cells, lining the lumen high assay variability associated with technical modifications of ducts and forming the alveoli) [11–13]. Subsequently, required for each cell population under study; these lim- Gudjonsson et al. [13], starting from the previous findings itations hamper the comparison of results obtained from [14], provided in vivo evidence of the morphogenic potential different studies and affect cell selection for in vitro and − + of the MUC-1 /ESA subpopulation, inoculating these cells in vivo growth experiments. In addition, recent findings subcutaneously in nude mice after pre-embedding them in a from two teratocarcinoma cell lines indicate that Hoechst mixture of collagen gel and matrigel. treatment, as performed during staining for SP analysis, can affect cell differentiation, suggesting other potential 2.2. Mammosphere Formation. To identify a human mam- complications in the interpretation of data [22]. mary stem/progenitor-cell subpopulation, Dontu et al. [15] Recently, another approach, based on aldehyde dehydro- adopted a strategy similar to that employed for primary genase (ALDH) activity, has been proposed as a promising neural cells and based on the formation of floating spherical alternative to identify and characterize the human mammary colonies used to define and measure stem cell-like behavior. stem/progenitor component in the mammary gland [23]. Journal of Oncology 3 CSC (a) Clonal evolution model (b) Cancer stem cell model Figure 1: Models of heterogeneity in solid cancer cells.(a) The clonal evolution model assumes that every cell in a tumor is potentially tumor- initiating. Progression is governed by rare stochastic events operating in all cells. Cells with mutations (yellow) that acquire growth advantage will dominate over all other cells in the tumor and will originate a new clone containing cells characterized by a different phenotype and having different proliferative potentials; in a clonogenicity or tumorigenicity assay, some of these cells (blue) would have a low probability of exhibiting this potential. (b) The cancer stem cell model states that a particular subset of tumor cells with stem cell-like properties, called “cancer stem cell” (CSC) (pink), drives tumor initiation, progression, and recurrence. CSCs are able to self-renew indefinitely and to differentiate, leading to the production of all cell types (blue) that make up the rest of the tumor. In clonogenic assays, CSCs have the potential to proliferate extensively and can form new tumors on transplantation. ALDH, a detoxifying enzyme responsible for the oxidation cell properties, with the consequence that human breast stem of intracellular aldehydes, is a putative candidate marker of cells have not yet been extensively characterized. stemness, since it is highly expressed in hematopoietic and neuronal stem/progenitor cells. Its presence can be evaluated 3. Cancer Propagation Models by ALDEFLUOR kit: brightly fluorescent ALDH-expressing cells are easily detected by flow cytometry in the green To explain why not every cell within a tumor is capable of fluorescence channel. Using such an approach, Ginestier et maintaining and/or reinitiating tumor growth, two models al. [23] showed that ALDH-positive cells formed mammo- of heterogeneity in solid cancer have been proposed: the spheres with high efficiency (10 ± 3.5%) when put into 96- clonal evolution model (Figure 1(a)) and the cancer stem cell well plates (1 cell/well), and displayed stem-like properties (CSC) model (Figure 1(b))[24–26]. in terms of bilineage differentiation in vitro and outgrowth The two main aspects of the clonal evolution model, first potential when inoculated in the mammary fat pad of proposed by Nowell in 1976 [27], are (1) diversity within humanized mice. However, even though the findings indicate the tumor due to genetic instability and (2) selection of the that only ALDH-positive cells had phenotypic and functional cells with the most advantageous phenotype. In this respect, characteristics of mammary stem cells, immunostaining of stem or differentiated cell characteristics (including self- tissue sections using a monoclonal antibody against the first renewing capacity) are just simple phenotypes and, as such, isoform of ALDH (ALDH1) did not detect any overlapping can change. According to this model, any cancer cell can expression of several markers (e.g., CK5/6 and CK14), pre- potentially become invasive and cause metastasis or become viously associated with undifferentiated mammary epithelial resistant to therapies and cause recurrence. cells, probably owing to the scarcity of this population. The cancer stem cell (CSC) model (Figure 1(b)) states that Analysis performed on mammosphere sections have shown a particular subset of tumor cells with stem cell-like proper- that ALDH1-positive cells represented approximately 5% of ties, called “cancer stem cells”, drives tumor initiation, pro- the total cell population and expressed CK5/6 or CK14, sup- gression, and recurrence. Since CSCs are widely believed to porting the hypothesis that ALDH1-positive cells represent arise from normal stem or progenitor cells, the identification the stem/progenitor population [13, 15]. of stem cells in a tissue is of paramount importance to Unfortunately, as highlighted by these inconclusive res- understand how a tumor arises. By definition, CSCs have ults, the efforts to purify adult stem cells from the human the ability to self-renew indefinitely and to differentiate, mammary gland have so far been hampered, on the one which leads to the production of all cell types composing hand, by the lack of cell surface markers specific to undiffer- a tumor, both tumorigenic and nontumorigenic cells. But entiated or differentiated mammary cells and, on the other the latter lack the unlimited self-renewing capacity and hand, by the lack of suitable in vivo assays for testing stem the ability to reproduce the phenotypically diverse cell 4 Journal of Oncology Clonal succession creating a dominant clone Mutation CSC2 CSC1 Figure 2: Mixed model for the nature of sustained tumor growth. The tumor is originally driven by rare cells of one phenotype (CSC1, yellow), which may have stem/progenitor cell origin and give rise to the tumor bulk by producing terminally differentiated cells (blue). Subsequent mutations enhancing self-renewing capacity create a dominant subclone that is phenotypically different (CSC2, green) and more aggressive. If the CSC2 subclone does not display “stem-like” cell properties, such a subset will not be able to initiate tumors with a high frequency [30]. populations that make up the tumor bulk [28, 29]. Therefore, pathways of stem cells. On the other hand, if these cells this model suggests that the presence of such rare tumor- arise from mature, differentiated cells, multiple oncogenic initiating cells, in the heterogeneous mix of cells composing a mutations, affecting differentiation and self-renewal path- tumor, is essential for neoplastic progression and metastatic ways are required for a cell to become tumorigenic and spread [29, 30]. While the CSC model of carcinogenesis was metastatic [43, 44]. It can be argued that mature cells have described in the context of systemic malignancies as long a very limited life span, and thus it is unlikely that all ago as in the 1930s, only recently has it been extended to the necessary mutations could occur during the relatively solid tumors, and during the last decade, several studies have short life of these cells. In contrast, the unlimited self- provided evidence that CSCs may also exist in solid tumors renewing capacity of normal stem cells could enable them including brain [31], lung [32], prostate [33], colon [34], to accumulate the necessary mutations, despite the apparent liver [35], pancreas [36], and breast [37] carcinomas, as well paradox of the stem cell dogma, according to which a as melanoma [38]. stem cell maintains its DNA constant through symmetric In breast tumorigenesis, the CSC model also seems to be division [26, 45]. Thus, whereas the CSC model is highly supported by clinical observations indicating that, despite hierarchical with a unique self-renewing cell type at the apex, the fact that breast cancer patients may have hundreds or the clonal evolution model attributes much of the intratumor thousands of single disseminated cancer cells detectable in variation to subclonal differences in mutational profile and their bloodstream, only a very small percentage of cells implies that all cells, except the terminally differentiated progresses to form overt macroscopic metastases [39], and ones, may have self-renewal capacity. Nevertheless, clonal that metastatic tumors tend to reproduce a heterogeneity evolution and CSC models share some aspects: in both, for similar to the primary tumor [40]. SinceCSCshavebeen instance, the tumor arises from a single cell that has acquired supposed to be responsible for the chemo- and radioresis- multiple mutations and has gained unlimited proliferative tance observed in several solid tumors [41, 42], the CSC potential. This suggests that tumor heterogeneity could be model could explain such a finding with the ability of CSCs to explained by a new version of the clonal evolution model escape cytotoxic drugs, via a high expression of specific drug that incorporates some features of the CSC hypothesis transporter proteins, and to resist radiotherapy by increasing [26]. DNA repair activity. A more intriguing model to explain the nature of sus- Unfortunately, the definitive proof of the cellular origin tained tumor growth is now emerging from the CSC model. of CSCs—they may arise either from normal resident stem According to this new model (Figure 2), tumors could origi- cells within the tissue bearing the malignancy or from nally be driven by rare CSCs (CSC1). Subsequent mutations, transformed progenitor cells that acquire the stem cell ability enhancing self-renewing capacity, could create a dominant, of self-renewal—remains elusive and is a topic of intense more aggressive subclone (CSC2), with a phenotypical aspect debate as well as experimental investigation. In fact, if distinct from the original CSC. However, if the CSC2 these cells arise from normal stem cells, then cancer cells subclone does not display “stem-like” properties, it should could take advantage of the existing regulatory self-renewal not be able to initiate tumors with a high frequency [46]. Journal of Oncology 5 CSC Progression CSC Mutation Blood mCSC vessels Invasion Metastasis Figure 3: Metastatic evolution. Genetic and epigenetic mechanisms may cause the generation of a self-renewing metastatic cancer stem cell subclone (mCSC, yellow), phenotypically different from the CSC that is driving tumorigenesis (pink). Through a series of invasive processes, mCSC enters blood vessels and colonizes distant organs according to the seed and soil hypothesis. Blue cells represent bulk tumor cells [47]. Although metastasis is the predominant cause of lethality 4. Isolation and Characterization of Breast in breast cancer patients, metastatic spread is a highly Cancer Stem Cells inefficient process, since very few cells successfully colonize On account of accumulating evidence and according to the distant sites. A possible explanation of this inefficiency is provided by the CSC model, according to which genetic CSC model that assumes a tissue stem/progenitor cell as the origin of a tumor, some methods adopted to isolate and epigenetic mechanisms may generate, within the pri- mary tumor, a self-renewing metastatic cancer stem cell and cultivate normal mammary stem/progenitor cells have (mCSC) characterized by an immunophenotype different also been applied to breast cancer tissue. However, there may be biases, inherent in the techniques applied, that can from the CSC that is driving tumorigenesis (Figure 3). This is suggested by the observation that in metastatic sites, affect experiment reproducibility. In fact, some technical approaches, including the strong enzymatic digestion of some cell subpopulations with self-renewing features display breast tissue necessary to disaggregate the connective tissue acellsurface marker profile different from the CSC that originated the primary tumor [47]. Through a series of surrounding the mammary gland, can cause damage to cells or loss of some particular surface markers, hamper- invasive processes, this new mCSC subclone could enter ing the identification of unique markers for the putative blood vessels and colonize distant organs according to the “seed and soil” hypothesis. CSCs (authors’ unpublished data). Furthermore, since cell 6 Journal of Oncology recovery from solid tissue is usually low (rarely exceeding metastatic and two primary invasive ductal carcinomas 10%), the samples obtained may not be representative of the (three triple negative tumors—ER-negative, PR-negative, original lesion, owing to the rare presence of CSCs in the and HER2-negative—and one ER-positive, PR-positive, and tumor mass, actual CSCs may be lost, while other cells may HER2-negative tumor) that were transplanted into the be mistakenly identified as CSCs [26]. humanized cleared fat pad of immunodeficient mice, imme- The first experimental clues about the existence of puta- diately after surgery with no previous cultivation. To test tive CSCs came from the observation that, even when using tumorigenicity, cell sorting was performed at early passages immortalized cancer cell lines, large numbers of cells (in the in animals, in order to minimize the variability introduced 5 6 range 10 –10 ) must be injected into experimental animals by the xenotransplant model. ALDEFLUOR-positive cells to initiate a tumor and, in spite of that, only a very small represented 3% to 10% of the total cell population, and proportion of these cells will go on to form metastases [48]. 500 positive cells were able to generate a tumor in as Additional experimental studies showed that while early few as 40 days. Significantly, the concomitant presence of steps in hematogenous metastasis (intravasation, survival, the ALDEFLUOR-positive phenotype and of the previously + −/low arrest, and extravasation) can be remarkably efficient, with described breast CSC phenotype (CD44 /CD24 )was over 80% of cells successfully completing the metastatic observed in a small cell fraction of the three triple-negative process at this point, only a small subset of these cells tumors (range 0.08–1.16%), whereas tumor cells generated (i.e., about 2%, depending on the experimental model) can from one metastatic tumor (pleural effusion) showed a initiate growth as micrometastases and that an even smaller high percentage of overlapping cell fraction (1.16%) that subset (i.e., about 0.02%, depending on the experimental gave rise to outgrowth from as few as 20 cells. Con- model) is able to persist and grow into macroscopic tumors. versely, ALDEFLUOR-negative cells, though bearing the + −/low This suggests that the initial growth of micrometastases CD44 /CD24 phenotype, were not tumorigenic, even represents the critical decision-making stage. when 50 000 cells/fat pad were implanted. This suggests that the concomitant presence of ALDEFLUOR-positive and + −/low CD44 /CD24 phenotypes may characterize progeni- 4.1. Surface Markers. After the identification of stemness tor cells with proliferative potential. When assessing the specific markers in hematopoietic tumors, considerable potential use of ALDH1 to detect malignant mammary progress has also been made in the elucidation of the stem/progenitor cells in situ on breast cancer tissue sections, biological properties of breast cancer stem cells. Al-Hajj et al. Ginestier et al. [23] found that ALDH1 expression correlated [37] first demonstrated the presence of a cell subpopulation with the histoclinical parameters, suggesting the use of this displaying stem cell properties, characterized by the cell marker as a powerful predictor of poor clinical outcome. + low surface marker profile CD44 /CD24 /lin , in solid tissues So far, therefore, the ALDEFLUOR assay, overcoming the and in pleural effusions of patients with advanced-stage limited availability of CSC-specific surface markers, seems to metastatic breast cancer. This phenotype displayed a 10- to represent the pivotal tool for the isolation of cell populations 50-fold increase in the ability to form tumors in NOD/SCID with high tumor-initiating capability or cell populations with mice over unfractionated tumor cells. In addition, the stem-like properties in normal tissue, allowing the identifica- authors demonstrated that cells with such a specific cell tion of stem/progenitor cells involved in normal mammary surface antigen profile could successfully and efficiently grow development and may be in tumor transformation. as tumor xenografts in immunodeficient mice; the highest As regards the use of Hoechst 33342 to detect the side capacity to form tumors was observed after injection of 200 population and to identify CSCs, thus bypassing the lack of + + low cells with the ESA /CD44 /CD24 /lin phenotype. universally accepted surface-antigen markers, several limi- However, the heterogeneous expression patterns of ESA, tations are emerging in addition to the technical criticisms CD44, or CD24, observed by FACS analysis in secondary described for the methods of normal mammary stem cell + low − lesions, support the hypothesis that the CD44 /CD24 /lin isolation; they include the low cell recovery from tumor profile could be the marker of the putative breast CSC phe- tissue that does not reflect the entire cohort of cancer cells, notype, since it recapitulates the heterogeneous complexity and the toxicity of the dye that precludes its use for functional of the tumors from which it has been isolated. Such a CSC assays in vitro and in vivo [50]. hypothesis could be corroborated by a study in which Ince et al. [49] observed the presence of two different populations 4.3. Self-Renewal Pathways. Since another trait shared by of mammary epithelial cells in the tumor of a single patient. normal stem cells and CSCs is the ability to self-renew, Only one population was myoepithelial-like and was able the deregulation of key pathways involved in such a pivotal to give rise to tumors with heterogeneous histology and to cellular function has been presumed to be implicated in be tumorigenic when injected into mammary fat pads of breast carcinogenesis and more thoroughly investigated. immunodeficient mice. Experimental evidence indicates that carcinogenesis in the mammary gland, and in other solid organs, might result in 4.2. Fluorescence Methods. Similar to normal breast stem the transformation of stem and/or progenitor cells because of cells, CSCs have also been investigated according to the deregulation of self-renewal pathways, including Notch, their expression of aldehyde dehydrogenase (ALDH) activ- Wnt, Hedgehog, and the transcription factor Bmi1 [51]and ity. Ginestier et al. [23] analyzed the tumorigenicity suggests that the targeting of self-renewal pathways might of ALDEFLUOR-positive populations isolated from two provide a specific approach to eradicate CSCs [52]. However, Journal of Oncology 7 in developing and testing compounds against putative CSCs, Meanwhile, the great challenge in stem cell investigation will several uncertainties must be faced and elucidated, first of all be the standardization of an orthotopic model in which the possible instability of CSCs that could hamper specific the whole tumor bulk could arise from a single definitively cell targeting. characterized human breast CSC. 5. Limits to the Xenograft Approach 6. Issues Concerning Established Breast Cancer Cell Lines There are several limitations to the use of xenograft assays as proof of “stemness”. The most relevant is that tumor growth Despite the intriguing results so far obtained, the use of and stem cell phenotype are not only determined by intrinsic established breast cancer cell lines as experimental models characteristics of tumor cells but are also influenced by the to collect data regarding CSCs is not without pitfalls, the microenvironment in which cells grow. In this respect, the main one being the attempt to apply stem-cell concepts to heterogeneity found in animal experimental models does breast cell lines. Established cell lines, in fact, are cultivated in not prove that normal or malignant stem cells undergo artificial conditions (depending on the experimental model) asymmetric division but just reflects a change in cell surface for many generations, with the risk that the unavoidable antigen expression induced by environmental conditions. selection induced by the serum media blurs the distinc- Simply injecting tumor cells into mice without measuring tion between tumorigenic and nontumorigenic clones. For the time of latency required to form a palpable tumor example, in vitro culture conditions, such as growth with or mass may induce to draw wrong conclusions about their without serum medium, could contribute to the functional absolute tumorigenic potential, which may be influenced differences found between non-CSCs and CSCs, including a by environmental conditions. Another important limitation diverse proliferative activity. to CSC investigation is the efficiency of thein vivo model It is unlikely that the so-called “cancer stem cells” derived used. It is well known that xenograft is less efficient than from established cell lines are the stem cells that make up syngeneic transplant because of the presence of animal a tumor. It is more likely that the “stem cell component” growth factors that could interact with their equivalent indicated as responsible for maintaining the line is in reality human receptors and provide confounding stimuli for the only a subpopulation of cells having a high-proliferative rate transplanted human stem/progenitor cells. In a recent paper, and being able to form clonal aggregates in the presence of Kelly et al. [53], challenging the CSC hypothesis, proposed additional techniques, for example, retroviral marking [16]. that xenograft may select a dominant clone capable of Therefore, extending the CSC concept to cell lines could surviving and maintaining tumor outgrowth in a foreign be misleading as is the case with the results reported by environment, and stressed the importance of performing a Sheridan et al. [56]. In a series of established breast cancer tumorigenic assay using cells sorted from the patient’s tumor cell lines (MDA-MB-231, TMD-436, Hs578T, SUM1315, to avoid cell variability and the selection of a dominant HBL-100, and MDA-MB-468), the authors observed a high cell subpopulation, after several serial passages in animals, + −/low percentage (>30%) of CD44 /CD24 cells which were whereas the majority of cells die owing to the lack of highly efficient in initiating a tumor in experimental settings, appropriate supporting factors. and concluded that those cell lines were composed mainly of As regards the mammary gland, in order to study human + −/low cancer stem cells. However, since the CD44 /CD24 phe- breast carcinogenesis, it is central to establish a model notype has been shown to be insufficient to confer stem-like system that more accurately recapitulates normal breast + −/low properties [23], such an enrichment in CD44 /CD24 epithelial development in rodents. For cancer cells, such a phenotype in established cell lines could be hypothesized system should also correlate with the clinical behavior of the to be a purification marker without functional implica- source tumor in patients. However, the inability of human tions. Therefore, it is crucial to demonstrate that such a breast epithelial cells to colonize mouse mammary fat pads subpopulation, contained in established breast cancer cell represents a constant problem. lines and displaying a putative stem-like phenotype, has The importance of both species- and tissue-specific the real functional characteristics of CSCs: self-renewal and influences has been highlighted in the studies by Kuperwasser differentiation capability. Awaiting these confirmations as et al. [54], which indicated that, although outgrowths can well as more standardized protocols for the isolation and be generated in the murine humanized mammary fat pads, expansion of CSCs from tissue, the established breast cancer the repopulating frequency by normal breast stem cells cell line model still remains a useful experimental tool to remains relatively low. This suggests that the expression of test drugs, radioresistance, and antibodies against the surface some markers of the inoculated cells could be influenced by markers associated with the putative stem-like phenotype. circulating or locally produced animal-specific factors. For example, estrogen has been found to profoundly affect the growth of ER-negative breast cancer cells because circulating 7. Microenvironment and Stem Cell Niche mouse estrogens led to recruitment of bone marrow-derived stromal cells and promoted the growth of tumors in virgin Despite the extensive use of the in vivo model, in which mice [55]. human tumor cells are injected into immunodeficient mice, However, despite these limitations, xenograft models still significant challenges are pending in experimental settings represent an essential tool for in vivo carcinogenesis studies. with CSCs, mostly related to the biological and technical 8 Journal of Oncology complexities associated with identifying, quantifying, and at least three lineage-restricted cell types outside the stem longitudinally monitoring CSCs in a complex in vivo envi- cell zone has been identified in the adult human breast ronment. In particular, as already mentioned, xenograft may [28]. The most likely location of the niche in the mature fail to reveal possible tumor growth-sustaining cells because gland could be the ducts, where in situ analysis identified the heterologous microenvironment precludes essential a narrow region of quiescent cells that were stained for interactions with support cells. As recently shown in an putative stem cell markers including chondroitin sulphate, elegant study by Naumov et al. [57], many established K6a, CK15, and SSEA-4 [69]. However, the data published human tumor cell lines, which had previously been described so far do not exclude the possible existence of other niches as “nontumorigenic” (including the breast cancer cell line or models of niche. In particular, the model of the estrogen- MDA-MB-436), were potentially tumorigenic but dormant. driven stem cell niche consists of three different cell types: In fact, when these cells were injected into animals and the ER-positive sensor cells, the EGFR-positive stromal cells, allowed to grow for a much longer period of time than a and the ER-negative stem cells. All of them could remain normal tumor growth assay (up to 12 months), spontaneous quiescent until they are switched on by estrogens. In response tumors eventually began to develop after an initial dormancy to estrogens, the ER-positive sensor cells synthesize and period and a switch to an angiogenic phenotype. Therefore, secrete amphiregulin that activates EGFR-positive stromal this study provides a conceptual framework to approach the cells which in turn activate ER-negative stem cells [70], problem that a supportive microenvironment is required for although the identity of the stromal factors interacting with tumor outgrowth in CSCs-involving studies. the epithelial components of the stem cell niche remains In adults, normal stem cells reside in a physiologically to be revealed. In addition, the similarities between stem unique microenvironment called stem cell niche. This niche, cell niches in different tissues remain poorly understood, in mainly composed of fibroblasts and myoepithelial cells, particular whether tissue-specific stem cells can be regulated provides a physical anchoring site for stem cells via adhesion by stem cell niches in other organs or whether vice versa molecules linking the stem cells to the extracellular matrix. ectopic mesenchymal stem cells may colonize a breast niche Support cells act as a hub in orientating dividing stem cells and influence its behavior. Studies by Hochedlinger et al. to hold one daughter cell in the niche, while the other one [71] support the notion that a malignant genome can be exits the niche and undergoes transit amplification followed reprogrammed to exhibit a normal-like phenotype when by differentiation [58, 59]. Under normal physiological transferred into a new biological context, whereas Blanpain conditions, the niche provides fine control over cell prolifera- et al. [72] have reported that epithelial stem cells are able to tion, typically balancing proliferation and apoptosis through generate their own microenvironment. Tumor cells also have paracrine factors, so that the stem cell population remains the well-established ability to interact with their surrounding undifferentiated and maintains a constant size [60, 61]. The environment and to influence it profoundly; examples effectors mediating heterotypic cell interactions within the are neoangiogenesis, recruitment of immune cells, and niche include a number of soluble factors and cell surface modification of tissue architecture. All these findings have receptors. Interestingly, some of these molecules, including important and provocative implications for understanding Wnt, Notch, TGF-β, bone morphogenetic proteins (BMPs), metastatic growth in secondary sites. and others [62–65], are known to be involved in tumor devel- opment, and emerging data support the idea that a “cancer stem cell niche” could also exist, and that interactions with 8. CSCs and Metastasis such a tumor niche may sustain a self-renewing population of tumor cells [66]. Alterations affecting stromal cells, such as Since the majority of breast cancer deaths occur as a result of local modifications of tissue homeostasis induced by chronic metastatic disease rather than from the effects of the primary inflammation, have been shown to promote formation of tumor, one of the biggest challenges is the identification, epithelial tumors, and very recent papers have described the as early as possible, of patients harboring metastatic cells. generation of pluripotent stem cells from fibroblasts. This In fact, the persistence of disease at a low or undetectable evidence supports the possibility that cancer may arise from level (the so-called “minimal residual disease”) is a common just a few mutations in resident tissue stem/progenitor cells feature of breast cancer as supported by autoptic findings or even differentiated cells leading to a stem-like phenotype [73] as well as by the accumulating evidence that breast [67, 68]. Indeed, if signaling pathways are dysregulated, cancer patients, even with no indication of metastatic spread the niche may be converted into a microenvironment that by current clinical parameters, have individual tumor cells favors uncontrolled proliferation and expansion of an altered in their blood [74, 75]. Several studies have shown that stem cell population. Similarly, a cancer stem cell could be detection of isolated tumor cells in the bone marrow is hypothesized to remain dormant in a metastatic site until an independent prognostic factor. However, even though activated by abnormal signaling from the microenvironment approximately 30% of breast cancer patients may have [66]. micrometastatic disease in their bone marrow at the onset, Currently, evidence of an anatomically and/or physi- only 30–50% of them will go on to develop clinically evident ologically specialized environment that constitutes a true metastases within 5 years [76, 77]. CSC niche is scarce, and the identity of a stem cell niche The presence of such cells in the bone marrow is within the mouse mammary gland has not been defined. particularly interesting since bone represents one of the Conversely, a putative stem cell niche which gives rise to most common sites for breast cancer metastasis, together Journal of Oncology 9 with regional lymph nodes, lung, liver, and brain, all of CSCs may constitute metastasis precursor cells, as suggested which may represent putative nichesfor disseminating tumor by the detection of disseminated tumor cells with a breast cells according to the hypothesis that cancer cells can arrest CSC phenotype in the bone marrow of breast cancer patients and grow in favorite metastatic sites. This “seed and soil” [83], it is of paramount importance to develop reliable theory, first proposed in 1889 by Paget [78], predicts that diagnostic tools through the identification of additional and a cancer cell (the “seed”) can survive in and colonize only more specific markers for CSCs or even for niche cells, secondary sites (the “soil”) that produce growth factors although this could be a difficult task due to the complexity capable of significantly influencing cell behavior [79, 80], of niche composition [89]. Difficult but not impossible, and has largely withstood the test of time [81]. In the case since recently Calabrese et al. [90]wereabletovisualize of breast cancer, disseminated carcinoma cells are detectable brain CSCs surviving in a vascular niche that secretes factors in the bone marrow [82], and recent findings indicate that which promote their long-term growth and self-renewal. In most cancer cells found in the bone marrow have a breast addition, innovative technologies, using sensitive imaging CSC phenotype [83]. techniques, have recently permitted real-time monitoring However, recent experimental observations demon- of CSC presence and viability as well as analysis of the strated a direct involvement of the bone marrow-derived angiogenic switch [91]. However, although this imaging cells in the development of human epithelial tumors, technique can be extremely valuable in order to achieve suggesting that CSCs may originate from bone marrow- a greater understanding of the biology of CSCs and their derived cells [84]. Furthermore, a very recent paper from relationship with the stromal compartment, it cannot be Mylona et al. [85] indicated that, in clinical breast cancer used, at least at present, for intravital monitoring of such + −/low tissues, the CD44 /CD24 phenotype (i.e., the phenotype elusive cells in patients. Conversely, more clinically relevant experimentally associated with stemness) had no significant imaging techniques, such as high-resolution magnetic res- correlation with clinical outcome. This is in agreement with onance imaging (MRI) [92] and three-dimensional high- a previous paper by Abraham et al. [86] that found no frequency ultrasound [93], are currently being developed for + −/low correlation between CD44 /CD24 tumor cell prevalence the study of CSCs in preclinical models. These techniques and tumor progression, in terms of event-free and overall look promising for future clinical applications to determine survival. Conversely and quite surprisingly, the prevalence prognosis, monitor therapeutic efficacy, and possibly affect − + of CD44 /CD24 tumor cells was found to exert an therapies. unfavorable impact on patients’ relapse-free and overall In addition to its impact on diagnosis and prognosis, + −/low survival. This suggests that, although CD44 /CD24 the CSC hypothesis also has significant implications for the breast tumor cells may be highly efficient in initiating therapy of breast cancer. Since current therapies do not tumors in animal experimental models, in patients, these target the tumor-initiating cells effectively, as implied by cells could be associated with the development of distant the large number of patients who relapse after adjuvant metastasis, particularly bone metastases, rather than with chemotherapeutic and/or hormonal treatment [94], there + −/low clinical outcome. Therefore, CD44 /CD24 tumor cells is a need for therapeutic agents specifically directed against could be a subclass of tumorigenic cells characterized by a CSCs. These specific agents could be added to conventional great metastatic potential, maybe due to the role of CD44 as cytotoxic drugs, designed to kill actively dividing cells, and a homing receptor for distant tissue compartments. be able to eradicate the metastatic disease. Thus, they would turn cancer, if detected at early stage, into a curable disease limited to the primary organ. Although we still know too little about the molecular features distinguishing CSCs from 9. Clinical Implications of CSC Paradigm the bulk of tumor cells to develop a “smart drug”, the and Future Directions significant advances in the field indicate that CSCs could The hypothesis that only CSCs are capable of reinitiating soon represent a really useful target. growth to form metastases in distant sites has fundamental clinical implications in terms of both prognosis and therapy 10. Conclusions since it provides an explanation of the limits to many current breast cancer treatments [87, 88]. In fact, the main goal of Recent findings in breast biology have provided support the current therapeutic strategies is represented by the “gold for the CSC hypothesis, but researchers still face many standard” of tumor shrinkage. However, if a tumor is main- challenges. First, attention should be paid to the accuracy tained by a small subpopulation of CSCs that is constitutively of experimental methods for the isolation and propagation resistant to therapeutic agents, tumor shrinkage results in the of CSCs derived from clinical samples, with a particular selective killing of the more differentiated, “nontumorigenic” emphasis on cell culture environment (substrate, atmo- cells that make up the bulk of the tumor, while leaving sphere, and medium) that has a critical role in standardizing cancer stem cells viable and able to continue to maintain the culture conditions for breast cancer progression studies and/or reinitiate tumor growth on a metastatic site. Thus, [49]. Secondly, more accurate techniques should be used current therapies fail to account for potential molecular and for the sphere formation assay, so as to determine a self- proliferative differences in the various subpopulations of renewing capability sufficient to classify a cancer cell as a tumor cells, and may be ineffective on the more aggressive cancer stem cell, and to avoid conflicting results obtained and dangerous subgroup that constituted by CSCs. Since by different groups. Moreover, it is necessary to pursue the 10 Journal of Oncology clinical demonstration that CSCs can be used as a prog- [8] J. Taylor-Papadimitriou, E. B. Lane, and S. E. Chang, “Cell lineages and interactions in neoplastic expression in the nostic indicator of disease progression and to identify the human breast,” in Understanding Breast Cancer: Clinical and mechanisms by which CSCs escape conventional therapies Laboratory Concepts, M. A. Rich, J. C. Hager, and P. Furmanski, in order to develop new specific therapeutic approaches. Eds., pp. 215–246, Marcel Dekker, New York, NY, USA, 1983. Finally, there is the need to find the definitive evidence [9] M. Shackleton, F. Vaillant, K. J. Simpson, et al., “Generation of of the existence of CSCs and to identify the stroma- a functional mammary gland from a single stem cell,” Nature, related factors that influence the development and spread vol. 439, no. 7072, pp. 84–88, 2006. of CSCs. Since CSCs have not yet been fully defined, their [10] J. Regan and M. Smalley, “Prospective isolation and functional existence in breast cancer cannot conclusively be proven, analysis of stem and differentiated cells from the mouse and competitive hypotheses, which may explain some of the mammary gland,” Stem Cell Reviews, vol. 3, no. 2, pp. 124– puzzling features of certain tumor cell populations, should 136, 2007. be taken into account. The most exciting of these competitive [11] J. Stingl, C. J. Eaves, U. Kuusk, and J. T. Emerman, “Phenotypic hypotheses implies the reversible epithelial-to-mesenchymal and functional characterization in vitro of a multipotent transition, the developmental process in which epithelial epithelial cell present in the normal adult human breast,” cells acquire the migratory properties of mesenchymal cells. Differentiation, vol. 63, no. 4, pp. 201–213, 1998. As shown in a very recent paper [95], the induction of [12] P. S. Rudland, R. Barraclough, D. G. Fernig, and J. A. Smith, the epithelial-mesenchymal transition could stimulate breast “Mammary stem cells in normal development and cancer,” in cells to adopt characteristics of stem cells. This suggests that Stem Cell, C. S. Potten, Ed., pp. 147–232, Accademic Press, San CSCs are not distinct entities but rather tumor cells that Diego, Calif, USA, 1997. transiently acquire stem cell-like properties as a consequence [13] T. Gudjonsson, R. Villadsen, H. L. Nielsen, L. Rønnov-Jessen, M. J. Bissell, and O. W. Petersen, “Isolation, immortalization, of an epithelial-mesenchymal transition. Undoubtedly, such and characterization of a human breast epithelial cell line with a link between epithelial-mesenchymal transition and stem stem cell properties,” Genes & Development,vol. 16, no.6,pp. cell phenotype further fuels the debated question about 693–706, 2002. the existence of CSCs and holds a number of interesting [14] J. Yang, R. C. Guzman, N. Popnikolov, et al., “Phenotypic implications for the biology of epithelial cells, including the characterization of collagen gel embedded primary human possibility that the stem cells of certain epithelial organs such breast epithelial cells in athymic nude mice,” Cancer Letters, as mammary glands may acquire many of the attributes of vol. 81, no. 2, pp. 117–127, 1994. the mesenchymal cell state that confer them an increased [15] G. Dontu, W. M. Abdallah, J. M. Foley, et al., “In vitro prop- tumorigenic potential. agation and transcriptional profiling of human mammary The advent of new technologies, including gene expres- stem/progenitor cells,” Genes & Development, vol. 17, no. 10, sion profiling and proteomics, and the ability to apply them pp. 1253–1270, 2003. to smallnumbers of cellswillprobablyhelptosolve such [16] I. Singec, R. Knoth, R. P. Meyer, et al., “Defining the actual open questions. sensitivity and specificity of the neurosphere assay in stem cell biology,” Nature Methods, vol. 3, no. 10, pp. 801–806, 2006. [17] J. Stingl and C. Caldas, “Molecular heterogeneity of breast References carcinomas and the cancer stem cell hypothesis,” Nature Reviews Cancer, vol. 7, no. 10, pp. 791–799, 2007. [1] A. Jemal, R. Siegel, E. Ward, et al., “Cancer statistics, 2006,” [18] L. A. Doyle, W. Yang, L. V. Abruzzo, et al., “A multidrug CA: A Cancer Journal for Clinicians, vol. 56, no. 2, pp. 106–130, resistance transporter from human MCF-7 breast cancer cells,” Proceedings of the National Academy of Sciences of the United [2] A.F.Chambers, A. C. Groom, andI.C.MacDonald, States of America, vol. 95, no. 26, pp. 15665–15670, 1998. “Dissemination and growth of cancer cells in metastatic sites,” [19] M. Kim, H. Turnquist, J. Jackson, et al., “The multidrug resis- Nature Reviews Cancer, vol. 2, no. 8, pp. 563–572, 2002. tance transporter ABCG2 (breast cancer resistance protein 1) [3] F. A. Tavassoli, P. Devilee, and World Health Organization effluxes Hoechst 33342 and is overexpressed in hematopoietic Classification of Tumors, Pathology and Genetics of Tumours stem cells,” Clinical Cancer Research, vol. 8, no. 1, pp. 22–28, of the Breast and Female Genital Organs, IARC Press, Lyon, France, 2003. [20] A. J. Alvi, H. Clayton, C. Joshi, et al., “Functional and molecu- [4] M. Duffy, “Estrogen receptors: role in breast cancer,” Critical lar characterisation of mammary side population cells,” Breast Reviews in Clinical Laboratory Sciences, vol. 43, no. 4, pp. 325– Cancer Research, vol. 5, no. 1, pp. R1–R8, 2002. 347, 2006. [21] R. P. Hill, “Identifying cancer stem cells in solid tumors: case [5] R. Schiff,S.A.Massarweh,J.Shou, et al., “Advancedconcepts not proven,” Cancer Research, vol. 66, no. 4, pp. 1891–1896, in estrogen receptor biology and breast cancer endocrine resistance: implicated role of growth factor signaling and [22] D. Adamski, J.-F. Mayol, N. Platet, F. Berger, F. Her ´ odin, and estrogen receptor coregulators,” Cancer Chemotherapy and D. Wion, “Effects of Hoechst 33342 on C2C12 and PC12 cell Pharmacology, vol. 56, supplement 1, pp. 10–20, 2005. differentiation,” FEBS Letters, vol. 581, no. 16, pp. 3076–3080, [6] J. S. Ross and J. A. Fletcher, “The HER-2/neu oncogene in breast cancer: prognostic factor, predictive factor, and target [23] C. Ginestier, M. H. Hur, E. Charafe-Jauffret, et al., “ALDH1 for therapy,” Stem Cells, vol. 16, no. 6, pp. 413–428, 1998. is a marker of normal and malignant human mammary stem [7] L. E. Ailles and I. L. Weissman, “Cancer stem cells in solid cells and a predictor of poor clinical outcome,” Cell Stem Cell, tumors,” Current Opinion in Biotechnology,vol. 18, no.5,pp. vol. 1, no. 5, pp. 555–567, 2007. 460–466, 2007. Journal of Oncology 11 [24] T. Reya,S.J.Morrison, M. F. Clarke,and I. L. Weissman,“Stem [44] D. Hanahan and R. A. Weinberg, “The hallmarks of cancer,” cells, cancer, and cancer stem cells,” Nature, vol. 414, no. 6859, Cell, vol. 100, no. 1, pp. 57–70, 2000. pp. 105–111, 2001. [45] J. Marx, “Cancer research. Mutant stem cells may seed cancer,” [25] R. Bjerkvig, B. B. Tysnes, K. S. Aboody, J. Najbauer, and Science, vol. 301, no. 5638, pp. 1308–1310, 2003. A. J. A. Terzis, “The origin of the cancer stem cell: current [46] J. M. Adams and A. Strasser, “Is tumor growth sustained by controversies and new insights,” Nature Reviews Cancer, vol. rare cancer stem cells or dominant clones?” Cancer Research, 5, no. 11, pp. 899–904, 2005. vol. 68, no. 11, pp. 4018–4021, 2008. [26] L. L. Campbell and K. Polyak, “Breast tumor heterogeneity: [47] J. E. Visvader and G. J. Lindeman, “Cancer stem cells in solid cancer stem cells or clonal evolution?” Cell Cycle, vol. 6, no. tumours: accumulating evidence and unresolved questions,” 19, pp. 2332–2338, 2007. Nature Reviews Cancer, vol. 8, no. 10, pp. 755–768, 2008. [27] P. C. Nowell, “The clonal evolution of tumor cell populations,” [48] D. R. Welch, “Technical considerations for studying cancer Science, vol. 194, no. 4260, pp. 23–28, 1976. metastasis in vivo,” Clinical and Experimental Metastasis, vol. [28] R. Villadsen, A. J. Fridriksdottir, L. Rønnov-Jessen, et al., 15, no. 3, pp. 272–306, 1997. “Evidence for a stem cell hierarchy in the adult human breast,” [49] T. A. Ince, A. L. Richardson, G. W. Bell, et al., “Transformation Journal of Cell Biology, vol. 177, no. 1, pp. 87–101, 2007. of different human breast epithelial cell types leads to distinct [29] M. S. Wicha, S. Liu, and G. Dontu, “Cancer stem cells: an old tumor phenotypes,” Cancer Cell, vol. 12, no. 2, pp. 160–170, idea—a paradigm shift,” Cancer Research, vol. 66, no. 4, pp. 1883–1890, 2006. [50] A. Hadnagy, L. Gaboury, R. Beaulieu, and D. Balicki, “SP [30] R. Pardal, M. F. Clarke, and S. J. Morrison, “Applying the analysis may be used to identify cancer stem cell populations,” principles of stem-cell biology to cancer,” Nature Reviews Experimental Cell Research, vol. 312, no. 19, pp. 3701–3710, Cancer, vol. 3, no. 12, pp. 895–902, 2003. 2006. [31] S. K. Singh, I. D. Clarke, M. Terasaki, et al., “Identification of [51] S. Liu, G. Dontu, and M. S. Wicha, “Mammary stem cells, a cancer stem cell in human brain tumors,” Cancer Research, self-renewal pathways, and carcinogenesis,” Breast Cancer vol. 63, no. 18, pp. 5821–5828, 2003. Research, vol. 7, no. 3, pp. 86–95, 2005. [32] C. D. Peacock and D. N. Watkins, “Cancer stem cells and the [52] M. Kubo, M. Nakamura, A. Tasaki, et al., “Hedgehog signaling ontogeny of lung cancer,” Journal of Clinical Oncology, vol. 26, pathway is a new therapeutic target for patients with breast no. 17, pp. 2883–2889, 2008. cancer,” Cancer Research, vol. 64, no. 17, pp. 6071–6074, 2004. [33] N. J. Maitland and A. T. Collins, “Prostate cancer stem cells: [53] P. N. Kelly, A. Dakic, J. M. Adams, S. L. Nutt, and A. Strasser, anew target fortherapy,” Journal of Clinical Oncology, vol. 26, “Tumor growth need not be driven by rare cancer stem cells,” no. 17, pp. 2862–2870, 2008. Science, vol. 317, no. 5836, p. 337, 2007. [34] L. Ricci-Vitiani, D. G. Lombardi, E. Pilozzi, et al., “Identifi- [54] C. Kuperwasser, T. Chavarria, M. Wu, et al., “Reconstruction cation and expansion of human colon-cancer-initiating cells,” of functionally normal and malignant human breast tissues in Nature, vol. 445, no. 7123, pp. 111–115, 2007. mice,” Proceedings of the National Academy of Sciences of the United States of America, vol. 101, no. 14, pp. 4966–4971, 2004. [35] S. Sell and H. L. Leffert, “Liver cancer stem cells,” Journal of Clinical Oncology, vol. 26, no. 17, pp. 2800–2805, 2008. [55] P. B. Gupta, D. Proia, O. Cingoz, et al., “Systemic stromal [36] C. J. Lee, J. Dosch, and D. M. Simeone, “Pancreatic cancer stem effects of estrogen promote the growth of estrogen receptor- negative cancers,” Cancer Research, vol. 67, no. 5, pp. 2062– cells,” Journal of Clinical Oncology, vol. 26, no. 17, pp. 2806– 2812, 2008. 2071, 2007. [37] M. Al-Hajj, M. S. Wicha, A. Benito-Hernandez, S. J. Morrison, [56] C. Sheridan, H. Kishimoto, R. K. Fuchs, et al., + − “CD44 /CD24 -breast cancer cells exhibit enhanced invase and M. F. Clarke, “Prospective identification of tumorigenic breast cancer cells,” Proceedings of the National Academy of properties: an early step necessary for metastasis,” Breast Cancer Research, vol. 8, no. 5, article R59, pp. 1–13, 2006. Sciences of the United States of America, vol. 100, no. 7, pp. 3983–3988, 2003. [57] G. N. Naumov, E. Bender, D. Zurakowski, et al., “A model [38] M. R. Kamstrup, R. Gniadecki, and G. L. Skovgaard, “Putative of human tumor dormancy: an angiogenic switch from the nonangiogenic phenotype,” Journal of the National Cancer cancer stem cells in cutaneous malignancies,” Experimental Dermatology, vol. 16, no. 4, pp. 297–301, 2007. Institute, vol. 98, no. 5, pp. 316–325, 2006. [39] L. Weiss, “Metastatic inefficiency,” Advances in Cancer [58] L. Li and T. Xie, “Stem cell niche: structure and function,” Research, vol. 54, pp. 159–211, 1990. Annual Review of Cell and Developmental Biology, vol. 21, pp. 605–631, 2005. [40] T. Brabletz, A. Jung, S. Spaderna, F. Hlubek, and T. Kirchner, “Migrating cancer stem cells—an integrated concept of malig- [59] A. Spradling, D. Drummond-Barbosa, and T. Kai, “Stem cells nant tumour progression,” Nature Reviews Cancer, vol. 5, no. find their niche,” Nature, vol. 414, no. 6859, pp. 98–104, 2001. 9, pp. 744–749, 2005. [60] T. Tumbar, G. Guasch, V. Greco, et al., “Defining the epithelial [41] P. Ø. Sakariassen, H. Immervoll, and M. Chekenya, “Cancer stem cell niche in skin,” Science, vol. 303, no. 5656, pp. 359– stem cells as mediators of treatment resistance in brain 363, 2004. tumors: status and controversies,” Neoplasia, vol. 9, no. 11, pp. [61] J. Zhang, C. Niu, L. Ye, et al., “Identification of the 882–892, 2007. haematopoietic stem cell niche and control of the niche size,” [42] M. Baumann, M. Krause, and R. Hill, “Exploring the role of Nature, vol. 425, no. 6960, pp. 836–841, 2003. cancer stem cells in radioresistance,” Nature Reviews Cancer, [62] G. Turashvili,J.Bouchal,G.Burkadze, andZ.Kolar, “Wnt vol. 8, no. 7, pp. 545–554, 2008. signalling pathway in mammary gland development and [43] M. Al-Hajj and M. F. Clarke, “Self-renewal and solid tumor carcinogenesis,” Pathobiology, vol. 73, no. 5, pp. 213–223, stem cells,” Oncogene, vol. 23, no. 43, pp. 7274–7282, 2004. 2006. 12 Journal of Oncology [63] G. Dontu, K. W. Jackson, E. McNicholas, M. J. Kawamura, W. [81] I. J. Fidler, “Seed and soil revisited: contribution of the organ M. Abdallah, and M. S. Wicha, “Role of Notch signaling in microenvironment to cancer metastasis,” Surgical Oncology cell-fate determination of human mammary stem/progenitor Clinics of North America, vol. 10, no. 2, pp. 257–269, 2001. cells,” Breast Cancer Research, vol. 6, no. 6, pp. R605–R615, [82] P. A. Phadke,R.R.Mercer, J. F. Harms, et al., “Kinetics of metastatic breast cancer cell trafficking in bone,” Clinical [64] S. Liu, G. Dontu, I. D. Mantle, et al., “Hedgehog signaling and Cancer Research, vol. 12, no. 5, pp. 1431–1440, 2006. Bmi-1 regulate self-renewal of normal and malignant human [83] M. Balic, H. Lin, L. Young, et al., “Most early disseminated mammary stem cells,” Cancer Research, vol. 66, no. 12, pp. cancer cells detected in bone marrow of breast cancer patients 6063–6071, 2006. have a putative breast cancer stem cell phenotype,” Clinical [65] B. Bierie andH.L.Moses,“TGF-β and cancer,” Cytokine & Cancer Research, vol. 12, no. 19, pp. 5615–5621, 2006. Growth Factor Reviews, vol. 17, no. 1-2, pp. 29–40, 2006. [84] C. R. Cogle, N. D. Theise, D. Fu, et al., “Bone marrow [66] L. Li and W. B. Neaves, “Normal stem cells and cancer stem contributes to epithelial cancers in mice and humans as cells: the niche matters,” Cancer Research, vol. 66, no. 9, pp. developmental mimicry,” Stem Cells, vol. 25, no. 8, pp. 1881– 4553–4557, 2006. 1887, 2007. [67] K. Takahashi, K. Tanabe, M. Ohnuki, et al., “Induction of [85] E. Mylona, I. Giannopoulou, E. Fasomytakis, et al., pluripotent stem cells from adult human fibroblasts by defined “The clinicopathologic and prognostic significance of + − − + factors,” Cell, vol. 131, no. 5, pp. 861–872, 2007. CD44 /CD24 /low and CD44 /CD24 tumor cells in invasive breast carcinomas,” Human Pathology, vol. 39, no. 7, [68] M. Wernig, A. Meissner, R. Foreman, et al., “In vitro pp. 1096–1102, 2008. reprogramming of fibroblasts into a pluripotent ES-cell-like state,” Nature, vol. 448, no. 7151, pp. 318–324, 2007. [86] B. K. Abraham, P. Fritz, M. McClellan, P. Hauptvogel, M. + −/low Athelogou, and H. Brauch, “Prevalence of CD44 /CD24 [69] M. A. LaBarge, O. W. Petersen, and M. J. Bissell, “Of cells in breast cancer may not be associated with clinical microenvironments and mammary stem cells,” Stem Cell outcome but may favor distant metastasis,” Clinical Cancer Reviews, vol. 3, no. 2, pp. 137–146, 2007. Research, vol. 11, no. 3, pp. 1154–1159, 2005. [70] C. Brisken and S. Duss, “Stem cells and the stem cell niche [87] M. Al-Hajj, M. W. Becker, M. Wicha, I. Weissman, and M. in the breast: an integrated hormonal and developmental F. Clarke, “Therapeutic implications of cancer stem cells,” perspective,” Stem Cell Reviews, vol. 3, no. 2, pp. 147–156, Current Opinion in Genetics and Development,vol. 14, no.1, pp. 43–47, 2004. [71] K. Hochedlinger, R. Blelloch, C. Brennan, et al., “Reprogram- [88] M. Dean, “Cancer stem cells: redefining the paradigm of ming of a melanoma genome by nuclear transplantation,” cancer treatment strategies,” Molecular Interventions, vol. 6, Genes & Development, vol. 18, no. 15, pp. 1875–1885, 2004. no. 3, pp. 140–148, 2006. [72] C. Blanpain, W. E. Lowry, A. Geoghegan, L. Polak, and E. [89] J. B. Sneddon and Z. Werb, “Location, location, location: the Fuchs, “Self-renewal, multipotency, and the existence of two cancer stem cell niche,” Cell Stem Cell, vol. 1, no. 6, pp. 607– cell populations within an epithelial stem cell niche,” Cell, vol. 611, 2007. 118, no. 5, pp. 635–648, 2004. [90] C. Calabrese, H. Poppleton, M. Kocak, et al., “A perivascular [73] J. Folkman and R. Kalluri, “Cancer without disease,” Nature, niche for brain tumor stem cells,” Cancer Cell,vol. 11, no.1, vol. 427, no. 6977, p. 787, 2004. pp. 69–82, 2007. [74] S. Meng, D. Tripathy, E. P. Frenkel, et al., “Circulating tumor [91] L. S. Hart and W. S. El-Deiry, “Invincible, but not invisible: cells in patients with breast cancer dormancy,” Clinical Cancer imaging approaches toward in vivo detection of cancer stem Research, vol. 10, no. 24, pp. 8152–8162, 2004. cells,” Journal of Clinical Oncology, vol. 26, no. 17, pp. 2901– [75] A. Ring, I. E. Smith, and M. Dowsett, “Circulating tumour 2910, 2008. cells in breast cancer,” Lancet Oncology, vol. 5, no. 2, pp. 79– [92] C. Heyn, J. A. Ronald, L. T. Mackenzie, et al., “In vivo magnetic 88, 2004. resonance imaging of single cells in mouse brain with optical [76] G. Wiedswang, E. Borgen, R. Karesen, et al., “Detection of validation,” Magnetic Resonance in Medicine, vol. 55, no. 1, pp. isolated tumor cells in bone marrow is an independent prog- 23–29, 2006. nostic factor in breast cancer,” Journal of Clinical Oncology, vol. [93] K. C. Graham,L.A.Wirtzfeld,L.T.MacKenzie,etal., 21, no. 18, pp. 3469–3478, 2003. “Three-dimensional high-frequency ultrasound imaging for [77] G. Gebauer, T. Fehm, E. Merkle, E. P. Beck, N. Lang, and W. longitudinal evaluation of liver metastases in preclinical Jager ¨ , “Epithelial cells in bone marrow of breast cancer patients models,” Cancer Research, vol. 65, no. 12, pp. 5231–5237, 2005. at time of primary surgery: clinical outcome during long-term [94] Early Breast Cancer Trialists’ Collaborative Group (EBCTCG), follow-up,” Journal of Clinical Oncology, vol. 19, no. 16, pp. “Effects of chemotherapy and hormonal therapy for early 3669–3674, 2001. breast cancer on recurrence and 15-year survival: an overview [78] S. Paget, “The distribution of secondary growths in cancer of of the randomised trials,” The Lancet, vol. 365, no. 9472, pp. the breast,” The Lancet, vol. 133, no. 3421, pp. 571–573, 1889. 1687–1717, 2005. [79] J. A. Aguirre-Ghiso, L. Ossowski, and S. K. Rosenbaum, [95] S. A. Mani, W. Guo, M.-J. Liao, et al., “The epithelial- “Green fluorescent protein tagging of extracellular signal- mesenchymal transition generates cells with properties of stem regulated kinase and p38 pathways reveals novel dynamics of cells,” Cell, vol. 133, no. 4, pp. 704–715, 2008. pathway activation during primary and metastatic growth,” Cancer Research, vol. 64, no. 20, pp. 7336–7345, 2004. [80] J. Folkman, “Role of angiogenesis in tumor growth and metastasis,” Seminars in Oncology, vol. 29, no. 6, supplement 16, pp. 15–18, 2002. MEDIATORS of INFLAMMATION The Scientific Gastroenterology Journal of World Journal Research and Practice Diabetes Research Disease Markers Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 International Journal of Journal of Immunology Research Endocrinology Hindawi Publishing Corporation Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 Submit your manuscripts at http://www.hindawi.com BioMed PPAR Research Research International Hindawi Publishing Corporation Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 Journal of Obesity Evidence-Based Journal of Journal of Stem Cells Complementary and Ophthalmology International Alternative Medicine Oncology Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 Parkinson’s Disease Computational and Behavioural Mathematical Methods AIDS Oxidative Medicine and in Medicine Research and Treatment Cellular Longevity Neurology Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Oncology Hindawi Publishing Corporation

The Controversial Clinicobiological Role of Breast Cancer Stem Cells

Loading next page...
 
/lp/hindawi-publishing-corporation/the-controversial-clinicobiological-role-of-breast-cancer-stem-cells-vKXzYac1FE
Publisher
Hindawi Publishing Corporation
Copyright
Copyright © 2008 Claudia Casarsa et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
ISSN
1687-8450
eISSN
1687-8469
DOI
10.1155/2008/492643
Publisher site
See Article on Publisher Site

Abstract

Hindawi Publishing Corporation Journal of Oncology Volume 2008, Article ID 492643, 12 pages doi:10.1155/2008/492643 Review Article The Controversial Clinicobiological Role of Breast Cancer Stem Cells 1 2 1 Claudia Casarsa, Saro Oriana, and Danila Coradini Experimental Oncology Laboratory, Senology Center, Ambrosiana Clinic, Cesano Boscone, 20090 Milano, Italy Surgery Department, Senology Center, Ambrosiana Clinic, Cesano Boscone, 20090 Milano, Italy Correspondence should be addressed to Danila Coradini, danila.coradini@yahoo.it Received 27 July 2008; Revised 5 December 2008; Accepted 23 December 2008 Recommended by Meenhard Herlyn Breast cancer remains a leading cause of morbidity and mortality in women mainly because of the propensity of primary breast tumors to metastasize. Growing experimental evidence suggests that cancer stem cells (CSCs) may contribute to tumor progression and metastasis spread. However, despite the tremendous clinical potential of such cells and their possible therapeutic management, the real nature of CSCs remains to be elucidated. Starting from what is currently known about normal mammary stem/progenitor cells, to better define the cell that originates a tumor or is responsible for metastatic spread, this review will discuss experimental evidence of breast cancer stem cells and speculate about the clinical importance and implications of their evaluation. Copyright © 2008 Claudia Casarsa et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. 1. Introduction of the tumor cells, invasive breast cancers are categorized by the World Health Organization into 18 different histological Despite significant advances in diagnosis and clinical man- subtypes, each of them associated with a diverse clinical behavior. In addition to morphology, invasive breast cancers agement, breast cancer remains a leading cause of morbidity and mortality in women [1], mainly owing to the propensity can also be classified according to their proliferative potential of primary breast tumors to metastasize to regional and (evaluated, e.g., by Ki67 expression) or the presence of such distant sites such as lymph nodes, lung, liver, bone, and brain biological factors as hormone receptors (estrogen [ER] and [2]. When the primary tumor is detected and removed before progesterone [PgR]) or HER2/neu overexpression that are metastasis occurs, prognosis could be good and the chance currently used in clinical practice to predict the prognosis of disease-free survival is high. However, if cancer cells have and the response/resistance to cytotoxic and/or hormonal already begun to disseminate from the primary tumor and therapy [4–6]. Understanding the molecular causes of such a hetero- spread to other organs, current therapeutic strategies largely depend on the use of systemic cytotoxic drugs that frequently geneity is therefore of paramount importance not only result in severe side effects on the patient and, in many cases, for the development of new therapeutic approaches, but do not yield long-term success. also for a better knowledge of the biological bases of This clinical scenario is further complicated by the fact breast tumorigenesis and metastatic spread. To address these that invasive breast cancers exhibit a wide range of mor- questions, over the past decade, scientists have used inno- phological types, molecular profiles, and clinical behaviors. vative technical strategies and approached new intriguing Not only there is a large variation in the nature of cell types directions aimed to define the genetic and epigenetic profile between cancers, but even within a single tumor a significant of the single tumor and to better define the cell that originates heterogeneity in phenotype and genotype can be observed a tumor or is responsible for metastatic spread. In particular, [3]. Based on growth patterns and cytological characteristics investigators focused their attention on the hypothesis that 2 Journal of Oncology breast cancer may be a stem cell disease, arising from In fact, contrary to the dogma that epithelial cell survival tissue stem/progenitor cells or driven by cells with stem-cell is anchorage-dependent, single cell suspensions of human properties [7]. mammary epithelial cells, obtained by mechanic/enzymatic dissociation, surprisingly survived in suspension and gener- ated floating spherical colonies, termed nonadherent mam- 2. Isolation and Characterization of Mammary mospheres [15]. The mammospheres contained numerous Stem/Progenitor Cells undifferentiated cells that, once isolated from the cluster, were able to generate new multilineage colonies, when In the adult, cell loss, associated with the physiological tissue cultured under differentiating conditions, and, in 3D culture, turnover, is compensated by the activity of specific cells, to reconstitute a functional mammary gland. However, the termed stem cells, which are so defined by their ability intrinsic dynamics of such cytospheres, as conventionally to self-renew and to generate the entire repertoire of the assayed, introduces several confounders, as reported in a differentiated cells composing a given tissue. Since 1983, recent paper by Singec et al. [16] who underlined the need the existence of such stem cells has also been postulated to use more accurate conditions for assessing the clonality, in the mammary gland to explain the cellular dynamics number, and fate of stem cells. Although sphere formation underlying morphological changes throughout a woman’s may represent a useful culturing tool, it is not specific to life, particularly during and after pregnancy [8]. stem cell characterization; any dividing cell from virtually In general, the identification and purification of normal any tissue will form floating cell clusters, when cultured in a stem cells are difficult tasks because of the paucity of stem serum-free medium and on a nonadherent substrate, owing cells in the tissue of origin and to the lack of stemness- to a predominant intercellular adhesiveness. Spontaneous specific morphologic traits. Hence, animal models have been sphere fusion may occur in normal as well as in neoplastic used, and the murine model, in particular, has supplied sphere cultures. Furthermore, in agreement with Singec et al. relevant data to improve our understanding of the cell [16], we have observed that the mammospheres, supposedly biology of the mammary gland and to clarify the presence rich in multilineage progenitors, have a very short life span of stem/progenitor cell and differentiated cell compartments (about 3-4 weeks), making it difficult to define the cells in the mammary gland [9, 10]. However, the information composing them as real mammary stem cells, which are long- obtained from mice cannot be directly applied to humans lived by definition. On account of all these criticisms, this (as in the case of murine exclusive cell surface markers) or experimental approach, based on the ability of the supposed may not be promptly translated to humans, as known in mammary stem cells to generate clonal mammospheres the hematopoietic field, where an opposite surface-antigen under anchorage-independent culture conditions, has been + − profile was observed in humans (CD34 CD38 )withrespect − + defined “a surrogate stem cell assay” [17]. to mice (CD34 CD38 ). Several in vitro strategies have thus been developed to isolate and characterize human mammary 2.3. Fluorescence Methods. To overcome the challenge repre- stem/progenitor cells, based on the differential expression of sented by the limited availability of stemness markers and to some cell surface markers, the formation of mammospheres, take advantage of the ability of stem cells to extrude dyes, for and the use of fluorescent dyes. example, Hoechst-33342 DNA-binding dye or rhodamine because the overexpression of some membrane transporter 2.1. Cell Surface Markers. The first experimental evidence proteins, such as P-glycoproteins or breast cancer resistance of the existence of human mammary stem/progenitor cells proteins (BCRPs) [18, 19], fluorescent dyes have been used was obtained by the in vitro isolation of multipotent to identify and isolate by flow cytometry a small fraction of epithelial cells, from normal human adult breast, accord- cells supposed to be stem progenitors [20]. This cell fraction, ing to their different expression of MUC-1 glycoprotein, which amounts to around 0.2% of the total population, has CALLA/CD10, and epithelial-specific antigen (ESA). Using been called side population (SP). these markers, two epithelial-cell progenitor populations However, several criticisms to such a sorting technique were distinguished, corresponding to the two components have been raised in the recent years principally concerning of normal mammary gland (myoepithelial cells, forming the the toxicity of dyes used in the analysis [21] and the basal layer of ducts, and epithelial cells, lining the lumen high assay variability associated with technical modifications of ducts and forming the alveoli) [11–13]. Subsequently, required for each cell population under study; these lim- Gudjonsson et al. [13], starting from the previous findings itations hamper the comparison of results obtained from [14], provided in vivo evidence of the morphogenic potential different studies and affect cell selection for in vitro and − + of the MUC-1 /ESA subpopulation, inoculating these cells in vivo growth experiments. In addition, recent findings subcutaneously in nude mice after pre-embedding them in a from two teratocarcinoma cell lines indicate that Hoechst mixture of collagen gel and matrigel. treatment, as performed during staining for SP analysis, can affect cell differentiation, suggesting other potential 2.2. Mammosphere Formation. To identify a human mam- complications in the interpretation of data [22]. mary stem/progenitor-cell subpopulation, Dontu et al. [15] Recently, another approach, based on aldehyde dehydro- adopted a strategy similar to that employed for primary genase (ALDH) activity, has been proposed as a promising neural cells and based on the formation of floating spherical alternative to identify and characterize the human mammary colonies used to define and measure stem cell-like behavior. stem/progenitor component in the mammary gland [23]. Journal of Oncology 3 CSC (a) Clonal evolution model (b) Cancer stem cell model Figure 1: Models of heterogeneity in solid cancer cells.(a) The clonal evolution model assumes that every cell in a tumor is potentially tumor- initiating. Progression is governed by rare stochastic events operating in all cells. Cells with mutations (yellow) that acquire growth advantage will dominate over all other cells in the tumor and will originate a new clone containing cells characterized by a different phenotype and having different proliferative potentials; in a clonogenicity or tumorigenicity assay, some of these cells (blue) would have a low probability of exhibiting this potential. (b) The cancer stem cell model states that a particular subset of tumor cells with stem cell-like properties, called “cancer stem cell” (CSC) (pink), drives tumor initiation, progression, and recurrence. CSCs are able to self-renew indefinitely and to differentiate, leading to the production of all cell types (blue) that make up the rest of the tumor. In clonogenic assays, CSCs have the potential to proliferate extensively and can form new tumors on transplantation. ALDH, a detoxifying enzyme responsible for the oxidation cell properties, with the consequence that human breast stem of intracellular aldehydes, is a putative candidate marker of cells have not yet been extensively characterized. stemness, since it is highly expressed in hematopoietic and neuronal stem/progenitor cells. Its presence can be evaluated 3. Cancer Propagation Models by ALDEFLUOR kit: brightly fluorescent ALDH-expressing cells are easily detected by flow cytometry in the green To explain why not every cell within a tumor is capable of fluorescence channel. Using such an approach, Ginestier et maintaining and/or reinitiating tumor growth, two models al. [23] showed that ALDH-positive cells formed mammo- of heterogeneity in solid cancer have been proposed: the spheres with high efficiency (10 ± 3.5%) when put into 96- clonal evolution model (Figure 1(a)) and the cancer stem cell well plates (1 cell/well), and displayed stem-like properties (CSC) model (Figure 1(b))[24–26]. in terms of bilineage differentiation in vitro and outgrowth The two main aspects of the clonal evolution model, first potential when inoculated in the mammary fat pad of proposed by Nowell in 1976 [27], are (1) diversity within humanized mice. However, even though the findings indicate the tumor due to genetic instability and (2) selection of the that only ALDH-positive cells had phenotypic and functional cells with the most advantageous phenotype. In this respect, characteristics of mammary stem cells, immunostaining of stem or differentiated cell characteristics (including self- tissue sections using a monoclonal antibody against the first renewing capacity) are just simple phenotypes and, as such, isoform of ALDH (ALDH1) did not detect any overlapping can change. According to this model, any cancer cell can expression of several markers (e.g., CK5/6 and CK14), pre- potentially become invasive and cause metastasis or become viously associated with undifferentiated mammary epithelial resistant to therapies and cause recurrence. cells, probably owing to the scarcity of this population. The cancer stem cell (CSC) model (Figure 1(b)) states that Analysis performed on mammosphere sections have shown a particular subset of tumor cells with stem cell-like proper- that ALDH1-positive cells represented approximately 5% of ties, called “cancer stem cells”, drives tumor initiation, pro- the total cell population and expressed CK5/6 or CK14, sup- gression, and recurrence. Since CSCs are widely believed to porting the hypothesis that ALDH1-positive cells represent arise from normal stem or progenitor cells, the identification the stem/progenitor population [13, 15]. of stem cells in a tissue is of paramount importance to Unfortunately, as highlighted by these inconclusive res- understand how a tumor arises. By definition, CSCs have ults, the efforts to purify adult stem cells from the human the ability to self-renew indefinitely and to differentiate, mammary gland have so far been hampered, on the one which leads to the production of all cell types composing hand, by the lack of cell surface markers specific to undiffer- a tumor, both tumorigenic and nontumorigenic cells. But entiated or differentiated mammary cells and, on the other the latter lack the unlimited self-renewing capacity and hand, by the lack of suitable in vivo assays for testing stem the ability to reproduce the phenotypically diverse cell 4 Journal of Oncology Clonal succession creating a dominant clone Mutation CSC2 CSC1 Figure 2: Mixed model for the nature of sustained tumor growth. The tumor is originally driven by rare cells of one phenotype (CSC1, yellow), which may have stem/progenitor cell origin and give rise to the tumor bulk by producing terminally differentiated cells (blue). Subsequent mutations enhancing self-renewing capacity create a dominant subclone that is phenotypically different (CSC2, green) and more aggressive. If the CSC2 subclone does not display “stem-like” cell properties, such a subset will not be able to initiate tumors with a high frequency [30]. populations that make up the tumor bulk [28, 29]. Therefore, pathways of stem cells. On the other hand, if these cells this model suggests that the presence of such rare tumor- arise from mature, differentiated cells, multiple oncogenic initiating cells, in the heterogeneous mix of cells composing a mutations, affecting differentiation and self-renewal path- tumor, is essential for neoplastic progression and metastatic ways are required for a cell to become tumorigenic and spread [29, 30]. While the CSC model of carcinogenesis was metastatic [43, 44]. It can be argued that mature cells have described in the context of systemic malignancies as long a very limited life span, and thus it is unlikely that all ago as in the 1930s, only recently has it been extended to the necessary mutations could occur during the relatively solid tumors, and during the last decade, several studies have short life of these cells. In contrast, the unlimited self- provided evidence that CSCs may also exist in solid tumors renewing capacity of normal stem cells could enable them including brain [31], lung [32], prostate [33], colon [34], to accumulate the necessary mutations, despite the apparent liver [35], pancreas [36], and breast [37] carcinomas, as well paradox of the stem cell dogma, according to which a as melanoma [38]. stem cell maintains its DNA constant through symmetric In breast tumorigenesis, the CSC model also seems to be division [26, 45]. Thus, whereas the CSC model is highly supported by clinical observations indicating that, despite hierarchical with a unique self-renewing cell type at the apex, the fact that breast cancer patients may have hundreds or the clonal evolution model attributes much of the intratumor thousands of single disseminated cancer cells detectable in variation to subclonal differences in mutational profile and their bloodstream, only a very small percentage of cells implies that all cells, except the terminally differentiated progresses to form overt macroscopic metastases [39], and ones, may have self-renewal capacity. Nevertheless, clonal that metastatic tumors tend to reproduce a heterogeneity evolution and CSC models share some aspects: in both, for similar to the primary tumor [40]. SinceCSCshavebeen instance, the tumor arises from a single cell that has acquired supposed to be responsible for the chemo- and radioresis- multiple mutations and has gained unlimited proliferative tance observed in several solid tumors [41, 42], the CSC potential. This suggests that tumor heterogeneity could be model could explain such a finding with the ability of CSCs to explained by a new version of the clonal evolution model escape cytotoxic drugs, via a high expression of specific drug that incorporates some features of the CSC hypothesis transporter proteins, and to resist radiotherapy by increasing [26]. DNA repair activity. A more intriguing model to explain the nature of sus- Unfortunately, the definitive proof of the cellular origin tained tumor growth is now emerging from the CSC model. of CSCs—they may arise either from normal resident stem According to this new model (Figure 2), tumors could origi- cells within the tissue bearing the malignancy or from nally be driven by rare CSCs (CSC1). Subsequent mutations, transformed progenitor cells that acquire the stem cell ability enhancing self-renewing capacity, could create a dominant, of self-renewal—remains elusive and is a topic of intense more aggressive subclone (CSC2), with a phenotypical aspect debate as well as experimental investigation. In fact, if distinct from the original CSC. However, if the CSC2 these cells arise from normal stem cells, then cancer cells subclone does not display “stem-like” properties, it should could take advantage of the existing regulatory self-renewal not be able to initiate tumors with a high frequency [46]. Journal of Oncology 5 CSC Progression CSC Mutation Blood mCSC vessels Invasion Metastasis Figure 3: Metastatic evolution. Genetic and epigenetic mechanisms may cause the generation of a self-renewing metastatic cancer stem cell subclone (mCSC, yellow), phenotypically different from the CSC that is driving tumorigenesis (pink). Through a series of invasive processes, mCSC enters blood vessels and colonizes distant organs according to the seed and soil hypothesis. Blue cells represent bulk tumor cells [47]. Although metastasis is the predominant cause of lethality 4. Isolation and Characterization of Breast in breast cancer patients, metastatic spread is a highly Cancer Stem Cells inefficient process, since very few cells successfully colonize On account of accumulating evidence and according to the distant sites. A possible explanation of this inefficiency is provided by the CSC model, according to which genetic CSC model that assumes a tissue stem/progenitor cell as the origin of a tumor, some methods adopted to isolate and epigenetic mechanisms may generate, within the pri- mary tumor, a self-renewing metastatic cancer stem cell and cultivate normal mammary stem/progenitor cells have (mCSC) characterized by an immunophenotype different also been applied to breast cancer tissue. However, there may be biases, inherent in the techniques applied, that can from the CSC that is driving tumorigenesis (Figure 3). This is suggested by the observation that in metastatic sites, affect experiment reproducibility. In fact, some technical approaches, including the strong enzymatic digestion of some cell subpopulations with self-renewing features display breast tissue necessary to disaggregate the connective tissue acellsurface marker profile different from the CSC that originated the primary tumor [47]. Through a series of surrounding the mammary gland, can cause damage to cells or loss of some particular surface markers, hamper- invasive processes, this new mCSC subclone could enter ing the identification of unique markers for the putative blood vessels and colonize distant organs according to the “seed and soil” hypothesis. CSCs (authors’ unpublished data). Furthermore, since cell 6 Journal of Oncology recovery from solid tissue is usually low (rarely exceeding metastatic and two primary invasive ductal carcinomas 10%), the samples obtained may not be representative of the (three triple negative tumors—ER-negative, PR-negative, original lesion, owing to the rare presence of CSCs in the and HER2-negative—and one ER-positive, PR-positive, and tumor mass, actual CSCs may be lost, while other cells may HER2-negative tumor) that were transplanted into the be mistakenly identified as CSCs [26]. humanized cleared fat pad of immunodeficient mice, imme- The first experimental clues about the existence of puta- diately after surgery with no previous cultivation. To test tive CSCs came from the observation that, even when using tumorigenicity, cell sorting was performed at early passages immortalized cancer cell lines, large numbers of cells (in the in animals, in order to minimize the variability introduced 5 6 range 10 –10 ) must be injected into experimental animals by the xenotransplant model. ALDEFLUOR-positive cells to initiate a tumor and, in spite of that, only a very small represented 3% to 10% of the total cell population, and proportion of these cells will go on to form metastases [48]. 500 positive cells were able to generate a tumor in as Additional experimental studies showed that while early few as 40 days. Significantly, the concomitant presence of steps in hematogenous metastasis (intravasation, survival, the ALDEFLUOR-positive phenotype and of the previously + −/low arrest, and extravasation) can be remarkably efficient, with described breast CSC phenotype (CD44 /CD24 )was over 80% of cells successfully completing the metastatic observed in a small cell fraction of the three triple-negative process at this point, only a small subset of these cells tumors (range 0.08–1.16%), whereas tumor cells generated (i.e., about 2%, depending on the experimental model) can from one metastatic tumor (pleural effusion) showed a initiate growth as micrometastases and that an even smaller high percentage of overlapping cell fraction (1.16%) that subset (i.e., about 0.02%, depending on the experimental gave rise to outgrowth from as few as 20 cells. Con- model) is able to persist and grow into macroscopic tumors. versely, ALDEFLUOR-negative cells, though bearing the + −/low This suggests that the initial growth of micrometastases CD44 /CD24 phenotype, were not tumorigenic, even represents the critical decision-making stage. when 50 000 cells/fat pad were implanted. This suggests that the concomitant presence of ALDEFLUOR-positive and + −/low CD44 /CD24 phenotypes may characterize progeni- 4.1. Surface Markers. After the identification of stemness tor cells with proliferative potential. When assessing the specific markers in hematopoietic tumors, considerable potential use of ALDH1 to detect malignant mammary progress has also been made in the elucidation of the stem/progenitor cells in situ on breast cancer tissue sections, biological properties of breast cancer stem cells. Al-Hajj et al. Ginestier et al. [23] found that ALDH1 expression correlated [37] first demonstrated the presence of a cell subpopulation with the histoclinical parameters, suggesting the use of this displaying stem cell properties, characterized by the cell marker as a powerful predictor of poor clinical outcome. + low surface marker profile CD44 /CD24 /lin , in solid tissues So far, therefore, the ALDEFLUOR assay, overcoming the and in pleural effusions of patients with advanced-stage limited availability of CSC-specific surface markers, seems to metastatic breast cancer. This phenotype displayed a 10- to represent the pivotal tool for the isolation of cell populations 50-fold increase in the ability to form tumors in NOD/SCID with high tumor-initiating capability or cell populations with mice over unfractionated tumor cells. In addition, the stem-like properties in normal tissue, allowing the identifica- authors demonstrated that cells with such a specific cell tion of stem/progenitor cells involved in normal mammary surface antigen profile could successfully and efficiently grow development and may be in tumor transformation. as tumor xenografts in immunodeficient mice; the highest As regards the use of Hoechst 33342 to detect the side capacity to form tumors was observed after injection of 200 population and to identify CSCs, thus bypassing the lack of + + low cells with the ESA /CD44 /CD24 /lin phenotype. universally accepted surface-antigen markers, several limi- However, the heterogeneous expression patterns of ESA, tations are emerging in addition to the technical criticisms CD44, or CD24, observed by FACS analysis in secondary described for the methods of normal mammary stem cell + low − lesions, support the hypothesis that the CD44 /CD24 /lin isolation; they include the low cell recovery from tumor profile could be the marker of the putative breast CSC phe- tissue that does not reflect the entire cohort of cancer cells, notype, since it recapitulates the heterogeneous complexity and the toxicity of the dye that precludes its use for functional of the tumors from which it has been isolated. Such a CSC assays in vitro and in vivo [50]. hypothesis could be corroborated by a study in which Ince et al. [49] observed the presence of two different populations 4.3. Self-Renewal Pathways. Since another trait shared by of mammary epithelial cells in the tumor of a single patient. normal stem cells and CSCs is the ability to self-renew, Only one population was myoepithelial-like and was able the deregulation of key pathways involved in such a pivotal to give rise to tumors with heterogeneous histology and to cellular function has been presumed to be implicated in be tumorigenic when injected into mammary fat pads of breast carcinogenesis and more thoroughly investigated. immunodeficient mice. Experimental evidence indicates that carcinogenesis in the mammary gland, and in other solid organs, might result in 4.2. Fluorescence Methods. Similar to normal breast stem the transformation of stem and/or progenitor cells because of cells, CSCs have also been investigated according to the deregulation of self-renewal pathways, including Notch, their expression of aldehyde dehydrogenase (ALDH) activ- Wnt, Hedgehog, and the transcription factor Bmi1 [51]and ity. Ginestier et al. [23] analyzed the tumorigenicity suggests that the targeting of self-renewal pathways might of ALDEFLUOR-positive populations isolated from two provide a specific approach to eradicate CSCs [52]. However, Journal of Oncology 7 in developing and testing compounds against putative CSCs, Meanwhile, the great challenge in stem cell investigation will several uncertainties must be faced and elucidated, first of all be the standardization of an orthotopic model in which the possible instability of CSCs that could hamper specific the whole tumor bulk could arise from a single definitively cell targeting. characterized human breast CSC. 5. Limits to the Xenograft Approach 6. Issues Concerning Established Breast Cancer Cell Lines There are several limitations to the use of xenograft assays as proof of “stemness”. The most relevant is that tumor growth Despite the intriguing results so far obtained, the use of and stem cell phenotype are not only determined by intrinsic established breast cancer cell lines as experimental models characteristics of tumor cells but are also influenced by the to collect data regarding CSCs is not without pitfalls, the microenvironment in which cells grow. In this respect, the main one being the attempt to apply stem-cell concepts to heterogeneity found in animal experimental models does breast cell lines. Established cell lines, in fact, are cultivated in not prove that normal or malignant stem cells undergo artificial conditions (depending on the experimental model) asymmetric division but just reflects a change in cell surface for many generations, with the risk that the unavoidable antigen expression induced by environmental conditions. selection induced by the serum media blurs the distinc- Simply injecting tumor cells into mice without measuring tion between tumorigenic and nontumorigenic clones. For the time of latency required to form a palpable tumor example, in vitro culture conditions, such as growth with or mass may induce to draw wrong conclusions about their without serum medium, could contribute to the functional absolute tumorigenic potential, which may be influenced differences found between non-CSCs and CSCs, including a by environmental conditions. Another important limitation diverse proliferative activity. to CSC investigation is the efficiency of thein vivo model It is unlikely that the so-called “cancer stem cells” derived used. It is well known that xenograft is less efficient than from established cell lines are the stem cells that make up syngeneic transplant because of the presence of animal a tumor. It is more likely that the “stem cell component” growth factors that could interact with their equivalent indicated as responsible for maintaining the line is in reality human receptors and provide confounding stimuli for the only a subpopulation of cells having a high-proliferative rate transplanted human stem/progenitor cells. In a recent paper, and being able to form clonal aggregates in the presence of Kelly et al. [53], challenging the CSC hypothesis, proposed additional techniques, for example, retroviral marking [16]. that xenograft may select a dominant clone capable of Therefore, extending the CSC concept to cell lines could surviving and maintaining tumor outgrowth in a foreign be misleading as is the case with the results reported by environment, and stressed the importance of performing a Sheridan et al. [56]. In a series of established breast cancer tumorigenic assay using cells sorted from the patient’s tumor cell lines (MDA-MB-231, TMD-436, Hs578T, SUM1315, to avoid cell variability and the selection of a dominant HBL-100, and MDA-MB-468), the authors observed a high cell subpopulation, after several serial passages in animals, + −/low percentage (>30%) of CD44 /CD24 cells which were whereas the majority of cells die owing to the lack of highly efficient in initiating a tumor in experimental settings, appropriate supporting factors. and concluded that those cell lines were composed mainly of As regards the mammary gland, in order to study human + −/low cancer stem cells. However, since the CD44 /CD24 phe- breast carcinogenesis, it is central to establish a model notype has been shown to be insufficient to confer stem-like system that more accurately recapitulates normal breast + −/low properties [23], such an enrichment in CD44 /CD24 epithelial development in rodents. For cancer cells, such a phenotype in established cell lines could be hypothesized system should also correlate with the clinical behavior of the to be a purification marker without functional implica- source tumor in patients. However, the inability of human tions. Therefore, it is crucial to demonstrate that such a breast epithelial cells to colonize mouse mammary fat pads subpopulation, contained in established breast cancer cell represents a constant problem. lines and displaying a putative stem-like phenotype, has The importance of both species- and tissue-specific the real functional characteristics of CSCs: self-renewal and influences has been highlighted in the studies by Kuperwasser differentiation capability. Awaiting these confirmations as et al. [54], which indicated that, although outgrowths can well as more standardized protocols for the isolation and be generated in the murine humanized mammary fat pads, expansion of CSCs from tissue, the established breast cancer the repopulating frequency by normal breast stem cells cell line model still remains a useful experimental tool to remains relatively low. This suggests that the expression of test drugs, radioresistance, and antibodies against the surface some markers of the inoculated cells could be influenced by markers associated with the putative stem-like phenotype. circulating or locally produced animal-specific factors. For example, estrogen has been found to profoundly affect the growth of ER-negative breast cancer cells because circulating 7. Microenvironment and Stem Cell Niche mouse estrogens led to recruitment of bone marrow-derived stromal cells and promoted the growth of tumors in virgin Despite the extensive use of the in vivo model, in which mice [55]. human tumor cells are injected into immunodeficient mice, However, despite these limitations, xenograft models still significant challenges are pending in experimental settings represent an essential tool for in vivo carcinogenesis studies. with CSCs, mostly related to the biological and technical 8 Journal of Oncology complexities associated with identifying, quantifying, and at least three lineage-restricted cell types outside the stem longitudinally monitoring CSCs in a complex in vivo envi- cell zone has been identified in the adult human breast ronment. In particular, as already mentioned, xenograft may [28]. The most likely location of the niche in the mature fail to reveal possible tumor growth-sustaining cells because gland could be the ducts, where in situ analysis identified the heterologous microenvironment precludes essential a narrow region of quiescent cells that were stained for interactions with support cells. As recently shown in an putative stem cell markers including chondroitin sulphate, elegant study by Naumov et al. [57], many established K6a, CK15, and SSEA-4 [69]. However, the data published human tumor cell lines, which had previously been described so far do not exclude the possible existence of other niches as “nontumorigenic” (including the breast cancer cell line or models of niche. In particular, the model of the estrogen- MDA-MB-436), were potentially tumorigenic but dormant. driven stem cell niche consists of three different cell types: In fact, when these cells were injected into animals and the ER-positive sensor cells, the EGFR-positive stromal cells, allowed to grow for a much longer period of time than a and the ER-negative stem cells. All of them could remain normal tumor growth assay (up to 12 months), spontaneous quiescent until they are switched on by estrogens. In response tumors eventually began to develop after an initial dormancy to estrogens, the ER-positive sensor cells synthesize and period and a switch to an angiogenic phenotype. Therefore, secrete amphiregulin that activates EGFR-positive stromal this study provides a conceptual framework to approach the cells which in turn activate ER-negative stem cells [70], problem that a supportive microenvironment is required for although the identity of the stromal factors interacting with tumor outgrowth in CSCs-involving studies. the epithelial components of the stem cell niche remains In adults, normal stem cells reside in a physiologically to be revealed. In addition, the similarities between stem unique microenvironment called stem cell niche. This niche, cell niches in different tissues remain poorly understood, in mainly composed of fibroblasts and myoepithelial cells, particular whether tissue-specific stem cells can be regulated provides a physical anchoring site for stem cells via adhesion by stem cell niches in other organs or whether vice versa molecules linking the stem cells to the extracellular matrix. ectopic mesenchymal stem cells may colonize a breast niche Support cells act as a hub in orientating dividing stem cells and influence its behavior. Studies by Hochedlinger et al. to hold one daughter cell in the niche, while the other one [71] support the notion that a malignant genome can be exits the niche and undergoes transit amplification followed reprogrammed to exhibit a normal-like phenotype when by differentiation [58, 59]. Under normal physiological transferred into a new biological context, whereas Blanpain conditions, the niche provides fine control over cell prolifera- et al. [72] have reported that epithelial stem cells are able to tion, typically balancing proliferation and apoptosis through generate their own microenvironment. Tumor cells also have paracrine factors, so that the stem cell population remains the well-established ability to interact with their surrounding undifferentiated and maintains a constant size [60, 61]. The environment and to influence it profoundly; examples effectors mediating heterotypic cell interactions within the are neoangiogenesis, recruitment of immune cells, and niche include a number of soluble factors and cell surface modification of tissue architecture. All these findings have receptors. Interestingly, some of these molecules, including important and provocative implications for understanding Wnt, Notch, TGF-β, bone morphogenetic proteins (BMPs), metastatic growth in secondary sites. and others [62–65], are known to be involved in tumor devel- opment, and emerging data support the idea that a “cancer stem cell niche” could also exist, and that interactions with 8. CSCs and Metastasis such a tumor niche may sustain a self-renewing population of tumor cells [66]. Alterations affecting stromal cells, such as Since the majority of breast cancer deaths occur as a result of local modifications of tissue homeostasis induced by chronic metastatic disease rather than from the effects of the primary inflammation, have been shown to promote formation of tumor, one of the biggest challenges is the identification, epithelial tumors, and very recent papers have described the as early as possible, of patients harboring metastatic cells. generation of pluripotent stem cells from fibroblasts. This In fact, the persistence of disease at a low or undetectable evidence supports the possibility that cancer may arise from level (the so-called “minimal residual disease”) is a common just a few mutations in resident tissue stem/progenitor cells feature of breast cancer as supported by autoptic findings or even differentiated cells leading to a stem-like phenotype [73] as well as by the accumulating evidence that breast [67, 68]. Indeed, if signaling pathways are dysregulated, cancer patients, even with no indication of metastatic spread the niche may be converted into a microenvironment that by current clinical parameters, have individual tumor cells favors uncontrolled proliferation and expansion of an altered in their blood [74, 75]. Several studies have shown that stem cell population. Similarly, a cancer stem cell could be detection of isolated tumor cells in the bone marrow is hypothesized to remain dormant in a metastatic site until an independent prognostic factor. However, even though activated by abnormal signaling from the microenvironment approximately 30% of breast cancer patients may have [66]. micrometastatic disease in their bone marrow at the onset, Currently, evidence of an anatomically and/or physi- only 30–50% of them will go on to develop clinically evident ologically specialized environment that constitutes a true metastases within 5 years [76, 77]. CSC niche is scarce, and the identity of a stem cell niche The presence of such cells in the bone marrow is within the mouse mammary gland has not been defined. particularly interesting since bone represents one of the Conversely, a putative stem cell niche which gives rise to most common sites for breast cancer metastasis, together Journal of Oncology 9 with regional lymph nodes, lung, liver, and brain, all of CSCs may constitute metastasis precursor cells, as suggested which may represent putative nichesfor disseminating tumor by the detection of disseminated tumor cells with a breast cells according to the hypothesis that cancer cells can arrest CSC phenotype in the bone marrow of breast cancer patients and grow in favorite metastatic sites. This “seed and soil” [83], it is of paramount importance to develop reliable theory, first proposed in 1889 by Paget [78], predicts that diagnostic tools through the identification of additional and a cancer cell (the “seed”) can survive in and colonize only more specific markers for CSCs or even for niche cells, secondary sites (the “soil”) that produce growth factors although this could be a difficult task due to the complexity capable of significantly influencing cell behavior [79, 80], of niche composition [89]. Difficult but not impossible, and has largely withstood the test of time [81]. In the case since recently Calabrese et al. [90]wereabletovisualize of breast cancer, disseminated carcinoma cells are detectable brain CSCs surviving in a vascular niche that secretes factors in the bone marrow [82], and recent findings indicate that which promote their long-term growth and self-renewal. In most cancer cells found in the bone marrow have a breast addition, innovative technologies, using sensitive imaging CSC phenotype [83]. techniques, have recently permitted real-time monitoring However, recent experimental observations demon- of CSC presence and viability as well as analysis of the strated a direct involvement of the bone marrow-derived angiogenic switch [91]. However, although this imaging cells in the development of human epithelial tumors, technique can be extremely valuable in order to achieve suggesting that CSCs may originate from bone marrow- a greater understanding of the biology of CSCs and their derived cells [84]. Furthermore, a very recent paper from relationship with the stromal compartment, it cannot be Mylona et al. [85] indicated that, in clinical breast cancer used, at least at present, for intravital monitoring of such + −/low tissues, the CD44 /CD24 phenotype (i.e., the phenotype elusive cells in patients. Conversely, more clinically relevant experimentally associated with stemness) had no significant imaging techniques, such as high-resolution magnetic res- correlation with clinical outcome. This is in agreement with onance imaging (MRI) [92] and three-dimensional high- a previous paper by Abraham et al. [86] that found no frequency ultrasound [93], are currently being developed for + −/low correlation between CD44 /CD24 tumor cell prevalence the study of CSCs in preclinical models. These techniques and tumor progression, in terms of event-free and overall look promising for future clinical applications to determine survival. Conversely and quite surprisingly, the prevalence prognosis, monitor therapeutic efficacy, and possibly affect − + of CD44 /CD24 tumor cells was found to exert an therapies. unfavorable impact on patients’ relapse-free and overall In addition to its impact on diagnosis and prognosis, + −/low survival. This suggests that, although CD44 /CD24 the CSC hypothesis also has significant implications for the breast tumor cells may be highly efficient in initiating therapy of breast cancer. Since current therapies do not tumors in animal experimental models, in patients, these target the tumor-initiating cells effectively, as implied by cells could be associated with the development of distant the large number of patients who relapse after adjuvant metastasis, particularly bone metastases, rather than with chemotherapeutic and/or hormonal treatment [94], there + −/low clinical outcome. Therefore, CD44 /CD24 tumor cells is a need for therapeutic agents specifically directed against could be a subclass of tumorigenic cells characterized by a CSCs. These specific agents could be added to conventional great metastatic potential, maybe due to the role of CD44 as cytotoxic drugs, designed to kill actively dividing cells, and a homing receptor for distant tissue compartments. be able to eradicate the metastatic disease. Thus, they would turn cancer, if detected at early stage, into a curable disease limited to the primary organ. Although we still know too little about the molecular features distinguishing CSCs from 9. Clinical Implications of CSC Paradigm the bulk of tumor cells to develop a “smart drug”, the and Future Directions significant advances in the field indicate that CSCs could The hypothesis that only CSCs are capable of reinitiating soon represent a really useful target. growth to form metastases in distant sites has fundamental clinical implications in terms of both prognosis and therapy 10. Conclusions since it provides an explanation of the limits to many current breast cancer treatments [87, 88]. In fact, the main goal of Recent findings in breast biology have provided support the current therapeutic strategies is represented by the “gold for the CSC hypothesis, but researchers still face many standard” of tumor shrinkage. However, if a tumor is main- challenges. First, attention should be paid to the accuracy tained by a small subpopulation of CSCs that is constitutively of experimental methods for the isolation and propagation resistant to therapeutic agents, tumor shrinkage results in the of CSCs derived from clinical samples, with a particular selective killing of the more differentiated, “nontumorigenic” emphasis on cell culture environment (substrate, atmo- cells that make up the bulk of the tumor, while leaving sphere, and medium) that has a critical role in standardizing cancer stem cells viable and able to continue to maintain the culture conditions for breast cancer progression studies and/or reinitiate tumor growth on a metastatic site. Thus, [49]. Secondly, more accurate techniques should be used current therapies fail to account for potential molecular and for the sphere formation assay, so as to determine a self- proliferative differences in the various subpopulations of renewing capability sufficient to classify a cancer cell as a tumor cells, and may be ineffective on the more aggressive cancer stem cell, and to avoid conflicting results obtained and dangerous subgroup that constituted by CSCs. Since by different groups. Moreover, it is necessary to pursue the 10 Journal of Oncology clinical demonstration that CSCs can be used as a prog- [8] J. Taylor-Papadimitriou, E. B. Lane, and S. E. Chang, “Cell lineages and interactions in neoplastic expression in the nostic indicator of disease progression and to identify the human breast,” in Understanding Breast Cancer: Clinical and mechanisms by which CSCs escape conventional therapies Laboratory Concepts, M. A. Rich, J. C. Hager, and P. Furmanski, in order to develop new specific therapeutic approaches. Eds., pp. 215–246, Marcel Dekker, New York, NY, USA, 1983. Finally, there is the need to find the definitive evidence [9] M. Shackleton, F. Vaillant, K. J. Simpson, et al., “Generation of of the existence of CSCs and to identify the stroma- a functional mammary gland from a single stem cell,” Nature, related factors that influence the development and spread vol. 439, no. 7072, pp. 84–88, 2006. of CSCs. Since CSCs have not yet been fully defined, their [10] J. Regan and M. Smalley, “Prospective isolation and functional existence in breast cancer cannot conclusively be proven, analysis of stem and differentiated cells from the mouse and competitive hypotheses, which may explain some of the mammary gland,” Stem Cell Reviews, vol. 3, no. 2, pp. 124– puzzling features of certain tumor cell populations, should 136, 2007. be taken into account. The most exciting of these competitive [11] J. Stingl, C. J. Eaves, U. Kuusk, and J. T. Emerman, “Phenotypic hypotheses implies the reversible epithelial-to-mesenchymal and functional characterization in vitro of a multipotent transition, the developmental process in which epithelial epithelial cell present in the normal adult human breast,” cells acquire the migratory properties of mesenchymal cells. Differentiation, vol. 63, no. 4, pp. 201–213, 1998. As shown in a very recent paper [95], the induction of [12] P. S. Rudland, R. Barraclough, D. G. Fernig, and J. A. Smith, the epithelial-mesenchymal transition could stimulate breast “Mammary stem cells in normal development and cancer,” in cells to adopt characteristics of stem cells. This suggests that Stem Cell, C. S. Potten, Ed., pp. 147–232, Accademic Press, San CSCs are not distinct entities but rather tumor cells that Diego, Calif, USA, 1997. transiently acquire stem cell-like properties as a consequence [13] T. Gudjonsson, R. Villadsen, H. L. Nielsen, L. Rønnov-Jessen, M. J. Bissell, and O. W. Petersen, “Isolation, immortalization, of an epithelial-mesenchymal transition. Undoubtedly, such and characterization of a human breast epithelial cell line with a link between epithelial-mesenchymal transition and stem stem cell properties,” Genes & Development,vol. 16, no.6,pp. cell phenotype further fuels the debated question about 693–706, 2002. the existence of CSCs and holds a number of interesting [14] J. Yang, R. C. Guzman, N. Popnikolov, et al., “Phenotypic implications for the biology of epithelial cells, including the characterization of collagen gel embedded primary human possibility that the stem cells of certain epithelial organs such breast epithelial cells in athymic nude mice,” Cancer Letters, as mammary glands may acquire many of the attributes of vol. 81, no. 2, pp. 117–127, 1994. the mesenchymal cell state that confer them an increased [15] G. Dontu, W. M. Abdallah, J. M. Foley, et al., “In vitro prop- tumorigenic potential. agation and transcriptional profiling of human mammary The advent of new technologies, including gene expres- stem/progenitor cells,” Genes & Development, vol. 17, no. 10, sion profiling and proteomics, and the ability to apply them pp. 1253–1270, 2003. to smallnumbers of cellswillprobablyhelptosolve such [16] I. Singec, R. Knoth, R. P. Meyer, et al., “Defining the actual open questions. sensitivity and specificity of the neurosphere assay in stem cell biology,” Nature Methods, vol. 3, no. 10, pp. 801–806, 2006. [17] J. Stingl and C. Caldas, “Molecular heterogeneity of breast References carcinomas and the cancer stem cell hypothesis,” Nature Reviews Cancer, vol. 7, no. 10, pp. 791–799, 2007. [1] A. Jemal, R. Siegel, E. Ward, et al., “Cancer statistics, 2006,” [18] L. A. Doyle, W. Yang, L. V. Abruzzo, et al., “A multidrug CA: A Cancer Journal for Clinicians, vol. 56, no. 2, pp. 106–130, resistance transporter from human MCF-7 breast cancer cells,” Proceedings of the National Academy of Sciences of the United [2] A.F.Chambers, A. C. Groom, andI.C.MacDonald, States of America, vol. 95, no. 26, pp. 15665–15670, 1998. “Dissemination and growth of cancer cells in metastatic sites,” [19] M. Kim, H. Turnquist, J. Jackson, et al., “The multidrug resis- Nature Reviews Cancer, vol. 2, no. 8, pp. 563–572, 2002. tance transporter ABCG2 (breast cancer resistance protein 1) [3] F. A. Tavassoli, P. Devilee, and World Health Organization effluxes Hoechst 33342 and is overexpressed in hematopoietic Classification of Tumors, Pathology and Genetics of Tumours stem cells,” Clinical Cancer Research, vol. 8, no. 1, pp. 22–28, of the Breast and Female Genital Organs, IARC Press, Lyon, France, 2003. [20] A. J. Alvi, H. Clayton, C. Joshi, et al., “Functional and molecu- [4] M. Duffy, “Estrogen receptors: role in breast cancer,” Critical lar characterisation of mammary side population cells,” Breast Reviews in Clinical Laboratory Sciences, vol. 43, no. 4, pp. 325– Cancer Research, vol. 5, no. 1, pp. R1–R8, 2002. 347, 2006. [21] R. P. Hill, “Identifying cancer stem cells in solid tumors: case [5] R. Schiff,S.A.Massarweh,J.Shou, et al., “Advancedconcepts not proven,” Cancer Research, vol. 66, no. 4, pp. 1891–1896, in estrogen receptor biology and breast cancer endocrine resistance: implicated role of growth factor signaling and [22] D. Adamski, J.-F. Mayol, N. Platet, F. Berger, F. Her ´ odin, and estrogen receptor coregulators,” Cancer Chemotherapy and D. Wion, “Effects of Hoechst 33342 on C2C12 and PC12 cell Pharmacology, vol. 56, supplement 1, pp. 10–20, 2005. differentiation,” FEBS Letters, vol. 581, no. 16, pp. 3076–3080, [6] J. S. Ross and J. A. Fletcher, “The HER-2/neu oncogene in breast cancer: prognostic factor, predictive factor, and target [23] C. Ginestier, M. H. Hur, E. Charafe-Jauffret, et al., “ALDH1 for therapy,” Stem Cells, vol. 16, no. 6, pp. 413–428, 1998. is a marker of normal and malignant human mammary stem [7] L. E. Ailles and I. L. Weissman, “Cancer stem cells in solid cells and a predictor of poor clinical outcome,” Cell Stem Cell, tumors,” Current Opinion in Biotechnology,vol. 18, no.5,pp. vol. 1, no. 5, pp. 555–567, 2007. 460–466, 2007. Journal of Oncology 11 [24] T. Reya,S.J.Morrison, M. F. Clarke,and I. L. Weissman,“Stem [44] D. Hanahan and R. A. Weinberg, “The hallmarks of cancer,” cells, cancer, and cancer stem cells,” Nature, vol. 414, no. 6859, Cell, vol. 100, no. 1, pp. 57–70, 2000. pp. 105–111, 2001. [45] J. Marx, “Cancer research. Mutant stem cells may seed cancer,” [25] R. Bjerkvig, B. B. Tysnes, K. S. Aboody, J. Najbauer, and Science, vol. 301, no. 5638, pp. 1308–1310, 2003. A. J. A. Terzis, “The origin of the cancer stem cell: current [46] J. M. Adams and A. Strasser, “Is tumor growth sustained by controversies and new insights,” Nature Reviews Cancer, vol. rare cancer stem cells or dominant clones?” Cancer Research, 5, no. 11, pp. 899–904, 2005. vol. 68, no. 11, pp. 4018–4021, 2008. [26] L. L. Campbell and K. Polyak, “Breast tumor heterogeneity: [47] J. E. Visvader and G. J. Lindeman, “Cancer stem cells in solid cancer stem cells or clonal evolution?” Cell Cycle, vol. 6, no. tumours: accumulating evidence and unresolved questions,” 19, pp. 2332–2338, 2007. Nature Reviews Cancer, vol. 8, no. 10, pp. 755–768, 2008. [27] P. C. Nowell, “The clonal evolution of tumor cell populations,” [48] D. R. Welch, “Technical considerations for studying cancer Science, vol. 194, no. 4260, pp. 23–28, 1976. metastasis in vivo,” Clinical and Experimental Metastasis, vol. [28] R. Villadsen, A. J. Fridriksdottir, L. Rønnov-Jessen, et al., 15, no. 3, pp. 272–306, 1997. “Evidence for a stem cell hierarchy in the adult human breast,” [49] T. A. Ince, A. L. Richardson, G. W. Bell, et al., “Transformation Journal of Cell Biology, vol. 177, no. 1, pp. 87–101, 2007. of different human breast epithelial cell types leads to distinct [29] M. S. Wicha, S. Liu, and G. Dontu, “Cancer stem cells: an old tumor phenotypes,” Cancer Cell, vol. 12, no. 2, pp. 160–170, idea—a paradigm shift,” Cancer Research, vol. 66, no. 4, pp. 1883–1890, 2006. [50] A. Hadnagy, L. Gaboury, R. Beaulieu, and D. Balicki, “SP [30] R. Pardal, M. F. Clarke, and S. J. Morrison, “Applying the analysis may be used to identify cancer stem cell populations,” principles of stem-cell biology to cancer,” Nature Reviews Experimental Cell Research, vol. 312, no. 19, pp. 3701–3710, Cancer, vol. 3, no. 12, pp. 895–902, 2003. 2006. [31] S. K. Singh, I. D. Clarke, M. Terasaki, et al., “Identification of [51] S. Liu, G. Dontu, and M. S. Wicha, “Mammary stem cells, a cancer stem cell in human brain tumors,” Cancer Research, self-renewal pathways, and carcinogenesis,” Breast Cancer vol. 63, no. 18, pp. 5821–5828, 2003. Research, vol. 7, no. 3, pp. 86–95, 2005. [32] C. D. Peacock and D. N. Watkins, “Cancer stem cells and the [52] M. Kubo, M. Nakamura, A. Tasaki, et al., “Hedgehog signaling ontogeny of lung cancer,” Journal of Clinical Oncology, vol. 26, pathway is a new therapeutic target for patients with breast no. 17, pp. 2883–2889, 2008. cancer,” Cancer Research, vol. 64, no. 17, pp. 6071–6074, 2004. [33] N. J. Maitland and A. T. Collins, “Prostate cancer stem cells: [53] P. N. Kelly, A. Dakic, J. M. Adams, S. L. Nutt, and A. Strasser, anew target fortherapy,” Journal of Clinical Oncology, vol. 26, “Tumor growth need not be driven by rare cancer stem cells,” no. 17, pp. 2862–2870, 2008. Science, vol. 317, no. 5836, p. 337, 2007. [34] L. Ricci-Vitiani, D. G. Lombardi, E. Pilozzi, et al., “Identifi- [54] C. Kuperwasser, T. Chavarria, M. Wu, et al., “Reconstruction cation and expansion of human colon-cancer-initiating cells,” of functionally normal and malignant human breast tissues in Nature, vol. 445, no. 7123, pp. 111–115, 2007. mice,” Proceedings of the National Academy of Sciences of the United States of America, vol. 101, no. 14, pp. 4966–4971, 2004. [35] S. Sell and H. L. Leffert, “Liver cancer stem cells,” Journal of Clinical Oncology, vol. 26, no. 17, pp. 2800–2805, 2008. [55] P. B. Gupta, D. Proia, O. Cingoz, et al., “Systemic stromal [36] C. J. Lee, J. Dosch, and D. M. Simeone, “Pancreatic cancer stem effects of estrogen promote the growth of estrogen receptor- negative cancers,” Cancer Research, vol. 67, no. 5, pp. 2062– cells,” Journal of Clinical Oncology, vol. 26, no. 17, pp. 2806– 2812, 2008. 2071, 2007. [37] M. Al-Hajj, M. S. Wicha, A. Benito-Hernandez, S. J. Morrison, [56] C. Sheridan, H. Kishimoto, R. K. Fuchs, et al., + − “CD44 /CD24 -breast cancer cells exhibit enhanced invase and M. F. Clarke, “Prospective identification of tumorigenic breast cancer cells,” Proceedings of the National Academy of properties: an early step necessary for metastasis,” Breast Cancer Research, vol. 8, no. 5, article R59, pp. 1–13, 2006. Sciences of the United States of America, vol. 100, no. 7, pp. 3983–3988, 2003. [57] G. N. Naumov, E. Bender, D. Zurakowski, et al., “A model [38] M. R. Kamstrup, R. Gniadecki, and G. L. Skovgaard, “Putative of human tumor dormancy: an angiogenic switch from the nonangiogenic phenotype,” Journal of the National Cancer cancer stem cells in cutaneous malignancies,” Experimental Dermatology, vol. 16, no. 4, pp. 297–301, 2007. Institute, vol. 98, no. 5, pp. 316–325, 2006. [39] L. Weiss, “Metastatic inefficiency,” Advances in Cancer [58] L. Li and T. Xie, “Stem cell niche: structure and function,” Research, vol. 54, pp. 159–211, 1990. Annual Review of Cell and Developmental Biology, vol. 21, pp. 605–631, 2005. [40] T. Brabletz, A. Jung, S. Spaderna, F. Hlubek, and T. Kirchner, “Migrating cancer stem cells—an integrated concept of malig- [59] A. Spradling, D. Drummond-Barbosa, and T. Kai, “Stem cells nant tumour progression,” Nature Reviews Cancer, vol. 5, no. find their niche,” Nature, vol. 414, no. 6859, pp. 98–104, 2001. 9, pp. 744–749, 2005. [60] T. Tumbar, G. Guasch, V. Greco, et al., “Defining the epithelial [41] P. Ø. Sakariassen, H. Immervoll, and M. Chekenya, “Cancer stem cell niche in skin,” Science, vol. 303, no. 5656, pp. 359– stem cells as mediators of treatment resistance in brain 363, 2004. tumors: status and controversies,” Neoplasia, vol. 9, no. 11, pp. [61] J. Zhang, C. Niu, L. Ye, et al., “Identification of the 882–892, 2007. haematopoietic stem cell niche and control of the niche size,” [42] M. Baumann, M. Krause, and R. Hill, “Exploring the role of Nature, vol. 425, no. 6960, pp. 836–841, 2003. cancer stem cells in radioresistance,” Nature Reviews Cancer, [62] G. Turashvili,J.Bouchal,G.Burkadze, andZ.Kolar, “Wnt vol. 8, no. 7, pp. 545–554, 2008. signalling pathway in mammary gland development and [43] M. Al-Hajj and M. F. Clarke, “Self-renewal and solid tumor carcinogenesis,” Pathobiology, vol. 73, no. 5, pp. 213–223, stem cells,” Oncogene, vol. 23, no. 43, pp. 7274–7282, 2004. 2006. 12 Journal of Oncology [63] G. Dontu, K. W. Jackson, E. McNicholas, M. J. Kawamura, W. [81] I. J. Fidler, “Seed and soil revisited: contribution of the organ M. Abdallah, and M. S. Wicha, “Role of Notch signaling in microenvironment to cancer metastasis,” Surgical Oncology cell-fate determination of human mammary stem/progenitor Clinics of North America, vol. 10, no. 2, pp. 257–269, 2001. cells,” Breast Cancer Research, vol. 6, no. 6, pp. R605–R615, [82] P. A. Phadke,R.R.Mercer, J. F. Harms, et al., “Kinetics of metastatic breast cancer cell trafficking in bone,” Clinical [64] S. Liu, G. Dontu, I. D. Mantle, et al., “Hedgehog signaling and Cancer Research, vol. 12, no. 5, pp. 1431–1440, 2006. Bmi-1 regulate self-renewal of normal and malignant human [83] M. Balic, H. Lin, L. Young, et al., “Most early disseminated mammary stem cells,” Cancer Research, vol. 66, no. 12, pp. cancer cells detected in bone marrow of breast cancer patients 6063–6071, 2006. have a putative breast cancer stem cell phenotype,” Clinical [65] B. Bierie andH.L.Moses,“TGF-β and cancer,” Cytokine & Cancer Research, vol. 12, no. 19, pp. 5615–5621, 2006. Growth Factor Reviews, vol. 17, no. 1-2, pp. 29–40, 2006. [84] C. R. Cogle, N. D. Theise, D. Fu, et al., “Bone marrow [66] L. Li and W. B. Neaves, “Normal stem cells and cancer stem contributes to epithelial cancers in mice and humans as cells: the niche matters,” Cancer Research, vol. 66, no. 9, pp. developmental mimicry,” Stem Cells, vol. 25, no. 8, pp. 1881– 4553–4557, 2006. 1887, 2007. [67] K. Takahashi, K. Tanabe, M. Ohnuki, et al., “Induction of [85] E. Mylona, I. Giannopoulou, E. Fasomytakis, et al., pluripotent stem cells from adult human fibroblasts by defined “The clinicopathologic and prognostic significance of + − − + factors,” Cell, vol. 131, no. 5, pp. 861–872, 2007. CD44 /CD24 /low and CD44 /CD24 tumor cells in invasive breast carcinomas,” Human Pathology, vol. 39, no. 7, [68] M. Wernig, A. Meissner, R. Foreman, et al., “In vitro pp. 1096–1102, 2008. reprogramming of fibroblasts into a pluripotent ES-cell-like state,” Nature, vol. 448, no. 7151, pp. 318–324, 2007. [86] B. K. Abraham, P. Fritz, M. McClellan, P. Hauptvogel, M. + −/low Athelogou, and H. Brauch, “Prevalence of CD44 /CD24 [69] M. A. LaBarge, O. W. Petersen, and M. J. Bissell, “Of cells in breast cancer may not be associated with clinical microenvironments and mammary stem cells,” Stem Cell outcome but may favor distant metastasis,” Clinical Cancer Reviews, vol. 3, no. 2, pp. 137–146, 2007. Research, vol. 11, no. 3, pp. 1154–1159, 2005. [70] C. Brisken and S. Duss, “Stem cells and the stem cell niche [87] M. Al-Hajj, M. W. Becker, M. Wicha, I. Weissman, and M. in the breast: an integrated hormonal and developmental F. Clarke, “Therapeutic implications of cancer stem cells,” perspective,” Stem Cell Reviews, vol. 3, no. 2, pp. 147–156, Current Opinion in Genetics and Development,vol. 14, no.1, pp. 43–47, 2004. [71] K. Hochedlinger, R. Blelloch, C. Brennan, et al., “Reprogram- [88] M. Dean, “Cancer stem cells: redefining the paradigm of ming of a melanoma genome by nuclear transplantation,” cancer treatment strategies,” Molecular Interventions, vol. 6, Genes & Development, vol. 18, no. 15, pp. 1875–1885, 2004. no. 3, pp. 140–148, 2006. [72] C. Blanpain, W. E. Lowry, A. Geoghegan, L. Polak, and E. [89] J. B. Sneddon and Z. Werb, “Location, location, location: the Fuchs, “Self-renewal, multipotency, and the existence of two cancer stem cell niche,” Cell Stem Cell, vol. 1, no. 6, pp. 607– cell populations within an epithelial stem cell niche,” Cell, vol. 611, 2007. 118, no. 5, pp. 635–648, 2004. [90] C. Calabrese, H. Poppleton, M. Kocak, et al., “A perivascular [73] J. Folkman and R. Kalluri, “Cancer without disease,” Nature, niche for brain tumor stem cells,” Cancer Cell,vol. 11, no.1, vol. 427, no. 6977, p. 787, 2004. pp. 69–82, 2007. [74] S. Meng, D. Tripathy, E. P. Frenkel, et al., “Circulating tumor [91] L. S. Hart and W. S. El-Deiry, “Invincible, but not invisible: cells in patients with breast cancer dormancy,” Clinical Cancer imaging approaches toward in vivo detection of cancer stem Research, vol. 10, no. 24, pp. 8152–8162, 2004. cells,” Journal of Clinical Oncology, vol. 26, no. 17, pp. 2901– [75] A. Ring, I. E. Smith, and M. Dowsett, “Circulating tumour 2910, 2008. cells in breast cancer,” Lancet Oncology, vol. 5, no. 2, pp. 79– [92] C. Heyn, J. A. Ronald, L. T. Mackenzie, et al., “In vivo magnetic 88, 2004. resonance imaging of single cells in mouse brain with optical [76] G. Wiedswang, E. Borgen, R. Karesen, et al., “Detection of validation,” Magnetic Resonance in Medicine, vol. 55, no. 1, pp. isolated tumor cells in bone marrow is an independent prog- 23–29, 2006. nostic factor in breast cancer,” Journal of Clinical Oncology, vol. [93] K. C. Graham,L.A.Wirtzfeld,L.T.MacKenzie,etal., 21, no. 18, pp. 3469–3478, 2003. “Three-dimensional high-frequency ultrasound imaging for [77] G. Gebauer, T. Fehm, E. Merkle, E. P. Beck, N. Lang, and W. longitudinal evaluation of liver metastases in preclinical Jager ¨ , “Epithelial cells in bone marrow of breast cancer patients models,” Cancer Research, vol. 65, no. 12, pp. 5231–5237, 2005. at time of primary surgery: clinical outcome during long-term [94] Early Breast Cancer Trialists’ Collaborative Group (EBCTCG), follow-up,” Journal of Clinical Oncology, vol. 19, no. 16, pp. “Effects of chemotherapy and hormonal therapy for early 3669–3674, 2001. breast cancer on recurrence and 15-year survival: an overview [78] S. Paget, “The distribution of secondary growths in cancer of of the randomised trials,” The Lancet, vol. 365, no. 9472, pp. the breast,” The Lancet, vol. 133, no. 3421, pp. 571–573, 1889. 1687–1717, 2005. [79] J. A. Aguirre-Ghiso, L. Ossowski, and S. K. Rosenbaum, [95] S. A. Mani, W. Guo, M.-J. Liao, et al., “The epithelial- “Green fluorescent protein tagging of extracellular signal- mesenchymal transition generates cells with properties of stem regulated kinase and p38 pathways reveals novel dynamics of cells,” Cell, vol. 133, no. 4, pp. 704–715, 2008. pathway activation during primary and metastatic growth,” Cancer Research, vol. 64, no. 20, pp. 7336–7345, 2004. [80] J. Folkman, “Role of angiogenesis in tumor growth and metastasis,” Seminars in Oncology, vol. 29, no. 6, supplement 16, pp. 15–18, 2002. MEDIATORS of INFLAMMATION The Scientific Gastroenterology Journal of World Journal Research and Practice Diabetes Research Disease Markers Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 International Journal of Journal of Immunology Research Endocrinology Hindawi Publishing Corporation Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 Submit your manuscripts at http://www.hindawi.com BioMed PPAR Research Research International Hindawi Publishing Corporation Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 Journal of Obesity Evidence-Based Journal of Journal of Stem Cells Complementary and Ophthalmology International Alternative Medicine Oncology Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 Parkinson’s Disease Computational and Behavioural Mathematical Methods AIDS Oxidative Medicine and in Medicine Research and Treatment Cellular Longevity Neurology Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014

Journal

Journal of OncologyHindawi Publishing Corporation

Published: Mar 19, 2009

References