Breast Cancer Stem Cells: Current Advances and Clinical Implications

  • Ming Luo
  • Shawn G. Clouthier
  • Yadwinder Deol
  • Suling Liu
  • Sunitha Nagrath
  • Ebrahim Azizi
  • Max S. WichaEmail author
Part of the Methods in Molecular Biology book series (MIMB, volume 1293)


There is substantial evidence that many cancers, including breast cancer, are driven by a population of cells that display stem cell properties. These cells, termed cancer stem cells (CSCs) or tumor initiating cells, not only drive tumor initiation and growth but also mediate tumor metastasis and therapeutic resistance. In this chapter, we summarize current advances in CSC research with a major focus on breast CSCs (BCSCs). We review the prevailing methods to isolate and characterize BCSCs and recent evidence documenting their cellular origins and phenotypic plasticity that enables them to transition between mesenchymal and epithelial-like states. We describe in vitro and clinical evidence that these cells mediate metastasis and treatment resistance in breast cancer, the development of novel strategies to isolate circulating tumor cells (CTCs) that contain CSCs and the use of patient-derived xenograft (PDX) models in preclinical breast cancer research. Lastly, we highlight several signaling pathways that regulate BCSC self-renewal and describe clinical implications of targeting these cells for breast cancer treatment. The development of strategies to effectively target BCSCs has the potential to significantly improve the outcomes for patients with breast cancer.

Key words

Cancer stem cells (CSCs) Breast cancer Circulating tumor cells (CTCs) Phenotypic plasticity Patient-derived xenograft (PDX) 

1 Introduction

The mammary epithelium is mainly composed of two distinct lineages: an inner layer of luminal epithelial cells and an outer layer of basal/myoepithelial cells. At puberty, the rudimentary mammary epithelial trees, under the influence of sex hormones including estrogen and progesterone, expand considerably to form an elaborate branching network terminating in small lobular structures called terminal ductal lobular units (TDLU, for women) [1, 2]. In late pregnancy, the epithelium compartment attains maximal development, and luminal epithelial cells in TDLU undergo robust proliferation and differentiation to form lobular-alveolar buds called secretory alveoli that produce milk during lactation [3]. After weaning, these lobular-alveolar structures regress with massive apoptosis, and the mammary glands remodel to a virgin-like state for the next pregnancy cycle [3, 4]. The dynamic features of the mammary gland strongly suggest the existence of stem-like cells with the mammary epithelium. Indeed, early transplantation studies using mouse mammary fat pad segments as well as retroviral labeling of mammary epithelial cells revealed that the mammary epithelium is maintained by a multipotent mammary stem cell compartment [5, 6]. Recent prospective isolation and characterization of mammary stem cells and progenitor populations in the mouse [7, 8, 9, 10] and human [11, 12, 13, 14, 15] mammary gland provide further evidence that the mammary epithelium is organized in a hierarchical manner, and that a single mammary stem cell (MaSC), which resides in the basal/myoepithelial layer, can functionally reconstitute a mammary gland by giving rise to differentiated progenies through various lineage-restricted progenitor cells [8, 9, 16].

Despite a longstanding relationship between tumorigenesis and stem cell biology, it is only recently that advances in stem cells biology have enabled direct testing of this hypothesis. Only recently, John Dick’s team provided the first evidence that human acute myeloid leukemia is organized as a hierarchy that originates from a primitive hematopoietic cell with self-renewal capacity [17]. Breast cancer, analogous to the hierarchical organization of the normal mammary epithelium, is also characterized as a highly heterogeneous disease. The heterogeneity of breast cancers is manifested by their classification into a number of distinct subtypes, each with a characteristic transcriptome and molecular expression signature [16]. There is now substantial evidence that breast cancer is hierarchically organized and driven by a small fraction of tumor cells that display stem cell properties [18, 19, 20]. This small population of breast cancer stem cells (BCSCs), also termed breast cancer initiating cells, were first identified among solid tumors by virtue of their expression of the cell surface markers EpCAM+, CD24, and CD44+ [21]. As few as 100 cells bearing this phenotype were able to produce tumors in immune deficient NOD/SCID mice, whereas over 100-fold greater cells that did not bear this phenotype was non-tumorigenic. Furthermore, tumors generated from EpCAM+CD24CD44+ BCSCs recapitulated the cell type heterogeneity of the initial tumor with a small fraction of tumor cells express the same stem-like signature, while the majority of bulk tumor cells do not. More recently, it has been shown that both normal and malignant breast stem cells express high level of the enzyme aldehyde dehydrogenase (ALDH) as assessed by the Aldefluor assay (Stem Cell Technologies, Inc., Vancouver, British Columbia, Canada). Moreover, ALDH1 expression serves as a predictor of poor clinical outcome in breast cancers [22]. Interestingly, the EpCAM+CD24CD44+ and ALDH+ populations across different subtypes of breast cancers identify anatomically distinct BCSCs with respective EMT (epithelial-to-mesenchymal transition) and MET (mesenchymal-to-epithelial transition) gene expression profiles, and they dynamically transition between the mesenchymal and epithelial states reflective of their normal counterparts in the mammary epithelial hierarchy [23].

The cancer stem cell (CSC) hypothesis and the prospective isolation and identification of a small fraction of highly tumorigenic cells with stem cell properties in leukemia, breast cancers and subsequently from a wide variety of other malignancies including that of the brain [24], prostate [25, 26], colon [27, 28], pancreas [29], liver [30, 31], lung [32], and head and neck [33] in the past decade has led to a paradigm shift in terms of how cancers form, progress and relapse. In this chapter, we provide a historical review and current state of knowledge on CSCs with a major focus on BCSCs, especially regarding the prevailing methods to isolate and characterize these cells, their potential cellular origin and phenotypic plasticity, and the role of BCSCs in promoting metastasis and treatment resistance of breast cancer. We also highlight the use of circulating tumor cells (CTCs) and patient-derived xenograft (PDX) models in preclinical breast cancer research, the main signaling pathways regulating BCSC self-renewal, and potential therapeutic approaches and clinical implications of targeting these cells for breast cancer treatment.

2 Stemness as a Concept in Cancer Biology

2.1 CSC Model as a Cellular Mechanism of Carcinogenesis and Tumor Heterogeneity

Although oncology researchers have only recently began to probe cancers for tumor stem cells, the concept of stem-driven carcinogenesis dates back to the middle of the nineteenth century [34]. The German pathologist Rudolf Virchow proposed in 1855 that cancer arises from newly activated embryonic-like cells resident in a once dormant state in mature tissue [34, 35, 36]. More than 150 years later direct evidence for the existence and importance of CSCs in tumor biology emerged. For years, clinicians and research scientists have recognized that the behavior of metastatic cancers is highly reminiscent of that of stem cells. However, CSCs could not be identified until specific cell surface protein markers became available for distinguishing a rare population of cells present in acute myeloid leukemia [17]. These CSCs resident in the peripheral blood of patients with leukemia were found to be capable of initiating leukemia—at approximately 1 in every 250,000 cells—when transferred into immune-compromised mice. Moreover, subsequent analysis of populations of leukemia-initiating cells from various subtypes of the disease indicated that these cells are primitive, in other words, they are most closely related to nascent hematopoietic blood-forming cells rather than to more differentiated, function-specific blood cells. The identification of leukemia-initiating cells subsequently fostered attempts to isolate and characterize CSCs in solid tumors. Particularly, during the past decade stem cell-like populations have been identified in benign and malignant tumors of the breast, colon, brain, pancreas, liver, prostate, head and neck, and many other tissues and in melanoma [21, 22, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 37, 38, 39, 40, 41]. The inability to identify reliable markers capable of reliably discriminating CSCs was a major impediment to progress within this field particularly when there is ample evidence that a presumptive CSC population exists for many tissue malignancies. In other cancers in which CSCs remain to be identified, researchers are beginning to investigate known stem cell markers in the various cancer cell populations in an effort to derive novel targeted therapeutics. Through this approach solid evidence supporting the role of CSCs in various cancers has materialized in recent years. These findings provide strong supporting rationale in favor of the hypothesis that a stem/early primitive progenitor cell gone awry could explain the molecular basis of tumorigenesis.

The reemergence and establishment of the CSC model as a cellular mechanism to probe tumor heterogeneity represents one of the major advances in cancer biology over the past decade. The CSC model of tumorigenesis is fundamentally different from the classic stochastic or clonal evolution model, which proposes that tumors arise via random mutation and clonal selection. According to the stochastic model, all cells within a tumor have equal possibility to acquire genetic mutations necessary to drive tumor initiation and growth. Random accumulation of genomic abnormities may lead to the generation of divergent, but coexisting tumor cell clones, which may account for intratumoral heterogeneity. However, this view of tumorigenesis is hard to reconcile with recent discoveries that only a small fraction of tumor cells harboring a “tumor initiating signature” can efficiently generate secondary tumors that recapitulate the same heterogeneous cell populations as the initial tumor. Distinct from the stochastic model, the CSC hypothesis postulates that, in the breast as well as other tissues, cancer arises from stem-like cells with dysregulated self-renewal capability. Based on this model, cancer is derived from cells harboring tumorigenic mutations and/or epigenetic modifications, which endow deregulated self-renewal capacity in these cells. Tumorigenic mutations and/or epigenetic modifications are more likely to occur and accumulate in tissue stem cells and/or their early progenitor cells due to their relatively long lifespan, leading to formation of CSCs. Alternatively, these genetic and epigenetic aberrations could hit more terminal differentiated cell populations, leading to a dedifferentiation process and endowing these cells with stem cell properties. In both scenarios, CSCs are at the top of the hierarchy to drive tumor initiation, growth, and progression. Like their normal counterparts, being able to self-renew and differentiate are two hallmarks of CSCs regardless of their cellular origins. The self-renewal process drives tumor initiation and growth, while differentiation helps to generate the bulk tumor cells and maintain tumor cell heterogeneity.

2.2 CSCs Are Not Necessarily Rare

Several studies have questioned the rarity of tumor cells with stem cell properties and tumor-initiating capacity as well as the xenotransplantation assay used to assess these cell populations. A recent publication by Sean Morrison’s group suggested that melanoma cells did not in fact follow a stem cell model [42]. Therein they discovered through the use of a highly immune-compromised NOD/SCID Interleukin-2 Receptor gamma chain deficient (IL-2Rγ−/−) mouse model that tumor-initiating cell frequency rose dramatically (up to 25 % versus one in a million when using conventional NOD/SCID mice). The IL-2Rγ−/− NOD/SCID (NSG) mice lack NK function and are less able to reject tumor xenografts than the parental NOD/SCID mice. Another study supported Morrison’s contention that melanoma does not arise from stem cells. Using the H3K4 demethylase JARID1B as a biomarker, Roesch and colleagues characterized a small subpopulation of slow-cycling melanoma cells that cycle with doubling times of more than 4 weeks within the rapidly proliferating main tumor population [43]. Knockdown of JARID1B led to an initial acceleration of tumor growth followed by exhaustion, suggesting that the JARID1B-positive subpopulation was essential for continuous tumor growth. Furthermore, JARID1B negative cells could become positive and a single melanoma cells irrespective of selection were found to be tumorigenic, suggesting that melanoma heterogeneity and maintenance are a dynamic process mediated by a temporarily distinct subpopulation [43]. However, this study used only established melanoma cell lines rather than primary tumors. Using primary melanoma cells as well as melanoma cell lines, Fang et al. showed that melanomas can contain a subpopulation of stem-like cells that contribute to heterogeneity and tumorigenesis [41]. Recently, Weissman’s lab further demonstrated that human neural crest nerve growth factor receptor CD271 enriched for tumor initiating cells and transplantation of isolated CD271+, but not CD271, melanoma cells into engrafted human skin or bone in T-, B-, and natural-killer-deficient Rag2 −/− γc −/− mice resulted in melanoma and metastasized [44]. Besides CD271 as a marker for melanoma CSCs, recently the expression of ALDH1 has been shown to enrich human melanoma stem cells [45] suggesting that tumorigenic cancer stem cells do exist in human melanomas. Additional CSC research will undoubtedly shed more light on this line of inquiry.

Similar to melanomas, CSCs may not necessarily constitute a minor component of the tumor in other tissue malignancies. Kelly et al. (2007) showed that, when lymphoma and leukemia of mouse origin are transplanted into histocompatible mice, a very high frequency (at least 1 in 10) of the tumor cells can seed tumor growth, suggesting that the low frequency of tumor-sustaining cells observed in xenotransplantation studies may reflect the limited ability of human tumor cells to adapt to growth in a foreign mouse environment [46]. In breast cancer, poorly differentiated and high-grade tumors displayed higher content of BCSCs than did low-grade well-differentiated tumors such as luminal breast cancers. In terms of ALDH marker expression, the basal breast carcinoma cell lines contain a higher content of Aldefluor-positive cells compared to luminal breast cancer cell lines [47]. Moreover, ALDH1 expression in human breast cancer tissues is associated with poor clinical outcome [22] and the content of ALDH1-positive cancer stem cells also predict engraftment of primary human breast tumors in NOD/SCID mice [48]. Using transcriptional profile of PKH26-label-retaining mammary epithelial cells (human MaSC signature), a recent study also supported the notion that the heterogeneous phenotypical and molecular traits of human breast cancers are a function of their CSC content [49]. Thus, in high-grade or advanced stages of breast tumors, the pool of BCSCs is greatly increased and they may not represent the minor population of cells found in the early-staged tumors.

3 Isolation and Identification of BCSCs

In order to study the basic attributes of CSCs, it is necessary to first identify, isolate, purify, and characterize these cells by employing techniques that enable one to distinguish them from the bulk tumor cell population. This objective remains extremely challenging because CSCs often constitute a small fraction of the total cell population of a tumor and may express the same cell surface markers as their more differentiated counterparts. A major challenge in the field today is to define specific CSC markers that are not expressed on differentiated bulk tumor cells. Nevertheless, the functional characteristics of CSCs including self-renewal capability and differentiation potential can be exploited by various in vitro and in vivo assays in order to identify and characterize this important cell population. Among the most useful techniques for identification and enrichment of BCSCs are side population (SP) dye exclusion, specific expression of cell surface markers, aldehyde dehydrogenase activity, tumorsphere culture and label-retention assays such as PKH staining. The putative BCSCs enriched by these techniques can be further validated functionally by the serial transplantation assay in vivo to evaluate their tumorigenic and self-renewal potential.

Dye exclusion relies on the observation that stem cells, as opposed to differentiated cells, have the ability to efflux lipophilic dyes including Rhodamine or Hoechst owing to their increased expression of the ATP-binding cassette transporter proteins such as ABCG2/BCRP1. This dye exclusion activity results in a so-called SP fraction that can be further quantitated by flow cytometry [50]. Several studies have demonstrated that the SP component is enriched in cells capable of recreating the whole cell population [51]. These observations suggest that the SP fraction may contain stem cells capable of recapitulating a tumor. To date, SP cells have been found in bone marrow, normal solid tissues, tumors, and various cancer cell lines [52, 53]. In human and mouse mammary epithelial cells, SP cells have been reported to contain stem/progenitor cell compartment [54, 55, 56, 57]. SP cells have also been identified in different subtypes of human breast cancer cells, which display a more tumorigenic phenotype than the cells do not readily efflux vital dyes [52, 58, 59, 60].

Another useful tool for identifying CSCs including BCSCs is flow cytometry utilizing expressed cell surface antigens. In breast and pancreatic cancers, the cell surface markers CD44 and CD24 are commonly used to define a population enriched in CSCs [21, 29, 61, 62]. The CD24CD44+ BCSCs cells possess the ability to self-renew and to differentiate when injected into the mammary fat pad of NOD/SCID mice, resulting in engrafted mammary tumor formation and progression [21]. Although the CD24CD44+ phenotypic markers have been extensively used to isolate BCSCs in different cancer subtypes especially basal-like tumors [63], a recent study showed that both CD24CD44+ and CD24+CD44+ cell populations in ERα–negative breast tumors are tumorigenic in murine xenograft models, and a third population of tumor-initiating cells enriched in CD44+CD49fhiCD133/2hi subset displays heightened tumorigenicity and self-renewal in vivo, and the capacity to give rise to functional and molecular heterogeneity [64]. The use of cell surface markers CD49f (α6 integrin) and CD29 (β1 intergrin) together with CD24 (heat stable antigen) or EpCAM (epithelial specific antigen), have been shown to enrich for mammary stem/progenitors in the mouse and human mammary gland [8, 9, 13]. In breast cancer, elevated CD49f expression is associated with reduced survival [65] and knockdown of its partner CD104 decreases in vivo tumorigenicity [66]. Thus, CD49f+, combined with CD44+CD24−/low expression signature, may better identify and enrich BCSCs in different breast cancer subtypes.

BCSCs can also be segregated from mixed cell populations using the Aldefluor assay which relies upon aldehyde dehydrogenase (ALDH) activity [67]. ALDH is an enzyme responsible for the oxidation of intracellular aldehydes; it plays a vital role in stem cell differentiation via retinoic acid metabolism [68, 69]. The commercially available Aldefluor kit (StemCell Technologies, Inc., Vancouver, British Columbia, Canada) contains a BODIPY-aminoacetaldehyde (BAAA) substrate labeled with a fluorochrome that is converted into BODIPY-aminoacetate (BAA) by ALDH catabolism [68]. Cells expressing high ALDH activity have brighter fluorescence and can be enumerated on a standard cytometer when DEAB (an inhibitor of the enzymatic reaction) is used as the isotype control for FACS analysis. The Aldefluor assay has been used for both murine and human stem/progenitor populations in many different cancers. Those cancers known to harbor an Aldefluor positive population with stem properties include acute myeloid leukemia [70], breast [22, 47], lung [71], colon [72, 73], head and neck [38], prostate [74], and others. ALDH immunohistochemistry on a number of isoforms can readily identify CSCs within formalin-fixed, paraffin-embedded tissues with the ALDH1-A1 and ALDH1-A3 isoforms being the most preferentially expressed [22]. The combination of Aldefluor positivity with other unique stem cell surface markers such as CD133+ and CD24CD44+ has been shown to further label and locate ovarian CSCs and BCSCs in their respective tumor tissue slides [39, 47].

BCSCs can subsequently be sorted and assayed for clonogenic potential in vitro and tumorigenicity in vivo by xenotransplantation using immune-compromised mice. The latter is the gold standard for assessing BCSC activity. Tumorsphere assays have also been frequently used to examine stem/progenitor activity in vitro for many cancers including breast cancer, glioblastoma and other tissue malignancies. Tumorspheres are generated by plating cells in a petri dish or flask under low attachment, serum-free conditions supplemented with B27, hydrocortisone, insulin, fibroblast growth factor (FGF), and epidermal growth factor (EGF) [75]. Those few cells that self-renew under this anchorage-independent condition divide and form tumorspheres. Notably, cells isolated from tumorspheres exhibit multi-lineage differentiation potential when given serum and extracellular matrix such as collagen [75].

A more recent method to characterize CSCs in vitro is the cell membrane label-retaining assay. This assay uses the PKH fluorescent dye series, which consist of a fluorophore attached to a peptide backbone that irreversibly binds to the lipid bilayer of cell membranes [76]. The PKH67/PKH26 dyes that irreversibly bound to the lipid bilayer on cell membranes get equally partitioned among daughter cells during cell division; thus the intensity decreases in an exponential manner with each round of cell cycle [77]. Stem cells including CSCs are usually maintained via an asymmetric self-renewal process, where a stem cell and a daughter cell are formed. Following asymmetric cell division, the stem cell enters a quiescent state, while the daughter cell (known as the transient amplifying cell) undergoes rapid proliferation and differentiation. As a result, the stem cell retains the PKH dye and can thus be identified and isolated by flow cytometry. This method has, for example, been used to label murine hematopoietic stem cells [78]; it has also been used to track the homing of short- and long-term repopulating cells in bone marrow [79] and to detect labeled cells by fluorescence microscopy [80]. The use of PKH dye label-retaining mammosphere assay has recently been used to identify both normal MaSCs and BCSCs [49, 81]. In these studies, normal or malignant mammary epithelial cells were stained with PKH26, grown in primary and secondary mammosphere culture, and subsequently dissociated and sorted for PKH high, low, or negative cell populations. These distinct cell populations were further tested for mammary epithelial repopulating or tumorigenic capacities, and as expected, only the PKH26high cell population was found to enrich for stem cell activities [49, 81]. This method is useful for labeling cells of interest, but these dyes can inadvertently be transferred to surrounding cells confounding results [82]. Therefore, the technique of PKH dye exclusion is most useful when combined with another stem cell characterization assays to validate stem cell phenotypes.

4 Cellular Origins and Phenotypic Plasticity of BCSCs

4.1 BCSCs Transition Between EMT- and MET-Like States

In primary breast xenografts, the CD24CD44+ and ALDH+ BCSCs identify minimally overlapping, largely separate cell populations, with each capable of initiating tumors in NOD/SCID mice [22]. However, whether these different phenotypic populations identify distinct or independent BCSCs in the tumor remains an interesting question. Recently, using gene expression profiling of ALDH+ and CD24CD44+ BCSCs isolated across different subtypes of human breast cancer tissues together with multi-marker immunofluorescence including CD24, CD44 and ALDH1, it was determined that CD24CD44+ and ALDH+ cell populations identify anatomically distinct BCSCs with distinct EMT and MET gene-expression profiles respectively. The EMT-like CD24CD44+ BCSCs are primarily quiescent and localized at the tumor invasive front, while the MET-like BCSCs expressing ALDH1 are recycling, proliferative cells located mainly in the central part of primary tumors [23]. Importantly, the epithelial and mesenchymal states of BCSCs are not static; instead they display a cellular plasticity allowing them to transit between the EMT and MET states [23]. The interconversion of BCSCs from the EMT-like (which is EpCAMCD49f+ that expresses the stem cell markers CD24CD44+) and MET-like (which is EpCAM+CD49f+ that expresses the CSC marker ALDH) state, as illustrated in Fig. 1, has been shown to be regulated by microRNA networks including EMT-inducing mir-9, mir-100, mir-221, and mir-155 as well as MET-inducing mir-200, mir-205, and mir-93 [83].
Fig. 1

Breast cancer stem cells can exist in at two interchangeable states, one EMT-like (EpCAM CD49f+ that expresses the stem cell markers CD44+ CD24) and a MET-like population (which is EpCAM+ CD49f+ and expresses the CSC marker ALDH). These stem cells states are interconvertible and are regulated by microRNAs

These studies, together with the findings that distinct microRNAs regulate the transition of BCSCs between the CD24CD44+ and ALDH+ phenotypes [83, 84], led to a hypothetical model (Fig. 2) highlighting the role of different BCSC states in tumor growth and metastasis. As illustrated in Fig. 2a, the CD24CD44+ BCSCs exist in an EMT-like state that are E-cadherin and EpCAM negative, Vimentin positive, and relatively quiescent, whereas the MET-like ALDH+ BCSCs are cycling, and E-cadherin and EpCAM positive, and Vimentin negative. The transition between these states is most likely dictated by the tumor microenvironment. For example, TGF-β generated in the tumor milieu can induce EMT by downregulation of epithelial markers such as E-cadherin and upregulation of EMT-inducing factors, such as Twist and Snail [85, 86, 87]. Both the inflammatory immune response [87] and the hypoxic tumor environment [88] have also been reported to induce EMT in cancers. On the other hand, BMP signaling has been reported to induce MET via induction of miR-205 and the miR-200 family of microRNAs that are key regulators of MET [89].
Fig. 2

(a) Breast cancer stem cells may exist in an EMT-like state (that are relatively quiescent, E cadherin and ESA negative, Vimentin positive and expresses the stem cell markers CD44+ CD24) or MET-like states (that are cycling, E cadherin and ESA positive, Vimentin negative and expresses the stem cell marker ALDH+). A small fraction of stem cells express both EMT and MET markers and these appear to be the most highly tumorigenic population of cancer stem cells. (b) BCSCs located inside the tumor mass exist chiefly as MET cells which are highly proliferative and ALDH+. In contrast, EMT CSCs are located at the tumor invasive front and are characterized by CD44+ CD24 markers. EMT cells are highly invasive and mediate metastasis. An intermediary CSC state (CD44+ CD24 ALDH+) that is highly proliferative with invasive potential also resides inside the tumor

The demonstration that BCSCs exist in two interchangeable epithelial and mesenchymal states also helps define a model of how BCSCs drive primary tumor growth and metastasis. As illustrated in Fig. 2b, in primary breast cancer, the CD24CD44+ mesenchymal-like BCSCs mediate the tumor invasion toward the basal membrane surrounding the tumor and into the blood, where they survive due to their intrinsic quiescence and anoikis resistance. After extravasation of the blood circulation system, these mesenchymal-like BCSCs form micro-metastasis in distant organs, where tumor microenvironment in the remote sites induces MET, which is required for BCSC self-renewal and generation of macro-metastasis. The plasticity of BCSCs from a quiescent mesenchymal state to a proliferative epithelial-like state plays a critical role for these cells to establish sizable metastatic nodules in distant organs. Indeed, there is increasing experimental evidence to suggest that such transition, termed colonization, is essential for development of successful macrometastasis [90]. For example, using a spontaneous squamous cell carcinoma mouse model, Tsai et al. recently demonstrated that turning off Twist1 to allow reversion of EMT is essential for disseminated tumor cells to proliferate and form metastases [91]. In another study, Ocaña et al. demonstrated that temporal loss of the EMT inducer Prrx1 is required for cancer cells to form lung macrometastasis [92]. In breast cancer, the content of CD24CD44+ BCSCs in the primary tumor correlates with increased risk of distant metastasis. However, distant metastases formed from these tumors frequently show a higher differentiation rate compared to the primary tumor, as manifested by increased expression of the luminal epithelial marker CD24 [11]. Similarly, in mouse model of breast cancer driven by MMTV-PyMT oncogene, it has been shown that CD90+ circulating tumor cells are responsible for lung metastasis, however, the proportion of CD90+ tumor cells decreases again in differentiating and growing metastatic nodules [93]. Furthermore, reexpression of miR-200 family members and subsequently epithelial differentiation was found to promote metastatic colonization of breast cancer cells [94]. Most recently, specific expression of Id1 gene in breast cancer cells undergone EMT has been shown to induce MET through antagonism of Twist1, and this phenotypic switching is required for metastatic colonization in the lung [95].

4.2 Cell Origins of BCSCs and Breast Tumorigenesis

Breast cancers are highly heterogeneous with several distinct subtypes characterized by their distinct histological and molecular expression features. These include luminal breast cancers that express steroid hormone receptors including estrogen receptor (ER) and progesterone receptor (PR); HER2 positive breast cancers characterized by amplification of the ErbB2 gene; and basal breast cancers that are triple negative (TN) for the expression of ER, PR, and HER2. Clinically, the classification of breast cancers into these major molecular subtypes greatly facilitates the use of treatment strategies and helps predict treatment response and prognostic relevance. Recent advancement in global gene expression profiling of breast cancers have further subdivided this heterogeneous disease into six different subtypes, including luminal A, luminal B, HER2 overexpressing, basal-like, claudin-low, and normal breast-like [96, 97, 98, 99]. The existence of these distinct breast cancer subtypes suggests that mammary oncogenesis may be derived from different cellular origins in the mammary epithelial hierarchy. Alternatively, different subtypes of breast cancers may arise from a common precursor in the mammary epithelium, but different oncogenic events play a major role in determining the distinct tumor subtype.

An intriguing finding from recent gene expression profiling is that, both the CD24CD44+ and ALDH+ BCSC subsets isolated in different subtypes of primary breast cancers exhibit a remarkable similarity in their patterns of gene expression, although whole-tumor gene expression profiles are distinct across different subtypes [23]. This finding, coupled with the knowledge that BCSCs transit between the EMT and MET states, suggests a potential common cell of origin for BCSCs across different breast cancer subtypes. Previous studies suggested a potential common cell origin for breast cancer cells that are resistant to chemotherapy [100, 101]. In these studies, gene expression pattern of primary tumors obtained from breast cancer patients after treatment with docetaxel showed a similar genetic profile, irrespective of their original histological phenotype [100]. Furthermore, it has been shown that molecular profile of breast tumors obtained after chemotherapy closely resembles the gene expression profile of untreated CD24CD44+ BCSCs [101]. This BCSC signature was found in TN and hormone sensitive breast cancers treated with neoadjuvant chemotherapy and endocrine therapy respectively [101]. This study suggested that hormone-sensitive luminal breast cancers also contain a subpopulation of mesenchymal-like CD24CD44+ BCSCs, which are hormone-resistant and are enriched following neoadjuvant endocrine therapy.

The relationship between different BCSC populations and their corresponding normal counterparts in the mammary epithelial hierarchy remains controversial. In women carrying BRCA1 germ line mutations, the relative proportion of EpCAMCD49fhi basal epithelial cells that containing basal stem cells was shown to be decreased, whereas the proportion of EpCAM+CD49f+ luminal progenitor cells increased. This aberrant luminal progenitor cell population also showed increased clonogenic activity compared to non-BRCA1 mutation carriers. Moreover, genome-wide transcriptome analyses of different subtypes of breast cancers as well as different mammary epithelial subpopulations in human BRCA1 mutation carriers revealed that the luminal progenitor gene signature was most strongly associated with the basal subtype of breast cancer, while the basal/mammary stem cell signature was correlated to tumors classified as Normal-like and Claudin-low subtypes [13]. These studies, although suggested aberrant luminal progenitors as the cell origin of BRCA1 basal breast cancer, could not directly demonstrate a cell origin of BRCA1 mutation-associated basal breast cancers. Using a mouse model carrying conditional BRCA1 alleles, Molyneux et al. have further demonstrated that BRCA1 basal-like breast cancer originates from luminal epithelial progenitors but not from basal mammary stem cells [102]. In this study, specific deletion of BRCA1 in luminal progenitor cells by Blg-Cre led to mammary tumors that are predominantly high grade invasive ductal carcinomas of no special type (IDC-NST), which closely phenocopies human basal BRCA1 tumors. In contrast, cell-type specific deletion of BRCA1 in basal stem/myoepithelial compartment by K14-Cre generated tumors that are predominantly malignant adenomyoepitheliomas. Several other studies using mouse models and isolated mammary epithelial cell populations also indicated that basal stem cells and mature luminal epithelial cells are not important targets for tumor initiation, but rather luminal progenitor cells are emerging as key players in breast tumorigenesis [103, 104].

Although the aforementioned studies strongly suggest luminal progenitor cells as the origin of basal breast cancer, the relationship of between BCSCs and the stem/progenitor cells in the normal mammary gland remains to be defined. Recent lineage tracing studies of the mouse mammary gland suggested that distinct basal and luminal stem cells give rise to cells restricted to the basal and luminal lineage respectively under normal developmental conditions [105]. However, transplantation of the basal CD24medCD49fhi and luminal CD24hiCD49flow populations into cleared mammary fat pad of recipient mice induces their plasticity and enables either of the stem cell populations to regenerate an entire mammary epithelial tree [106]. The existence of distinct luminal and basal stem cells in the mouse mammary gland prompted us to evaluate if distinct luminal and basal stem cells with regenerative capability exist in the human mammary tissue, and if these stem cell populations are reflective of the ALDH+ and CD24CD44+ BCSCs isolated in different subtypes of breast cancers. Indeed, recent studies have demonstrated that the human mammary gland displays a similar hierarchy organization, containing functional luminal stem cells located in the terminal lobules that are EpCAM+CD49f+ALDH+, and basal stem cells located in the mammary ducts that are CD24CD44+EpCAMCD49f+ [23]. Interestingly, in the luminal compartment, only EpCAM+CD49f+ALDH+ luminal stem cells, which constitute 6 % of total EpCAM+CD49f+ population, have high colony-forming activity and generate ductal/alveolar structures in 3D Matrigel [23]. These studies suggest that ALDH serves as a functional marker of the luminal stem cells and the self-renewing/proliferative state of ALDH+ luminal stem cells may increase their susceptibility to carcinogenic mutations, making them a logical cell of origin for breast cancer. This notion was supported by studies showing that expansion of lobules containing ALDH1-expressing cells is associated with loss of heterozygosity in BRCA1 mutation carriers [107]. These studies also suggest that it is the luminal stem cells instead of luminal progenitors that are involved in the tumorigenesis of BRCA1 basal breast cancer.

The studies from Liu et al. together with recent observations of luminal progenitor cells as the key players in breast tumorigenesis led to a model of breast oncogenesis in which luminal stem cells serve as a common cell origin of breast cancers, but distinct oncogenic events dictate the individual tumor subtype. For example, loss of BRCA1 or BRCA1 mutation in the luminal stem cells promotes their tumorigeneis leading to the so-called basal breast cancer. However, further oncogenic events, such as loss of P53 and Pten, may drive EMT and promote the mesenchymal-like claudin-low phenotype. The EMT process could also be induced by the tumor microenvironment, leading to the mesenchymal-like claudin-low phenotype. On the other hand, oncogenic mutations such as the PIK3CA and HER2 amplification in luminal stem cells may promote the luminal and HER2 phenotype respectively.

5 BCSCs and Treatment Resistance

5.1 BCSCs Are Associated with Treatment Resistance

Current antitumor strategies are mainly aimed at shrinking tumor mass by targeting the rapidly proliferating bulk tumor cells. These therapeutic strategies, albeit effective in reducing the size of primary tumors, frequently fail to eradicate advanced tumors and are associated with tumor relapse. The CSC hypothesis suggests that conventional antitumor strategies targeting rapidly proliferating cells may fail to target CSCs that divide infrequently in the tumor. A plethora of studies have indicated that BCSCs are relatively resistant to both ionizing radiation and chemotherapy [108, 109, 110, 111] in cultured breast cancer cell lines, in primary mammary tumor cells derived from different mouse models of human breast cancer [112, 113, 114], and in patient derived tumor xenografts in mice [109, 115, 116].

The intrinsic resistance of BCSCs to neoadjuvant chemotherapy in the clinical setting has also been documented in a number of studies. Using primary breast cancer cells isolated from tumors treated with neoadjuvant chemotherapy and those resected from chemotherapy-naive patients, Yu and colleagues examined mammosphere forming activity as well as CD24CD44+ BCSC content in corresponding tumor samples. This study revealed a 14-fold increase in mammosphere forming activity (5.8 % vs. 0.4 %) and eightfold increase in CD24CD44+ enriched population (74 % vs. 9 %) after neoadjuvant chemotherapy [115]. The intrinsic resistance of BCSCs to neoadjuvant chemotherapy was also reported in paired breast cancer core biopsies obtained from 31 patients before and after 12 weeks of treatment with docetaxel or doxorubicin/cyclophosphamide [117]. This study revealed that the percentage of CD24CD44+ BCSCs and tumorsphere forming activity in the residual tumor tissues significantly increased following neoadjuvant chemotherapy. Interestingly, in a separate group of patients with HER2 amplification, treatment with HER2 and EGFR inhibitor Lapatinib following chemotherapy did not increase, but rather slightly decreased the content of CD24CD44+ BCSCs as well as tumorsphere forming activity [117]. Since HER2 overexpression has been show to drive BCSC activity [118], this clinical study suggests that strategies combining BCSC targeting agents (e.g., Trastuzumab and Lapatinib) with conventional chemotherapy hold the potential to overcome BCSC associated treatment resistance and achieve better therapeutic outcomes.

BCSCs can also be enriched by virtue of their expression of ALDH activity and ALDH1 expression in human breast tumors is associated with poor prognosis [22]. Consistent with this observations, a clinical study examining ALDH1 expression in a cohort of primary breast cancer samples treated with sequential paclitaxel and epirubicin-based chemotherapy revealed that ALDH1 positivity was significantly associated with a low pathological complete response (pCR) rate and resistance to the therapy [119]. This study suggests that ALDH1 positive BCSCs also share the properties of resistance to conventional chemotherapy.

5.2 Mechanisms Involving Treatment Resistance in BCSCs

5.2.1 BCSCs with Enhanced Drug Efflux Activity

One mechanism that has been described to explain the innate resistance of BCSCs to neoadjuvant chemotherapy is the elevated efflux activity found in BCSCs. In the mouse mammary gland, a population of long-term BrdU label retaining cells has been shown to enrich for stem cell antigen-1 (Sca-1) and Hoechst dye-effluxing “side population” properties associated with increased regenerative potential when transplanted into the cleared mammary fat pads of host mice [57]. SP cells with stem/progenitor activity have also been identified in human mammary tissue which constitute 0.2–5.0 % of the total cell population [54, 55, 56]. Transcriptional profiling of mammary gland SP cells has revealed an upregulation of genes associated with multidrug resistance [120]. BCSCs, like the normal mammary stem/progenitor cells, may share this “side population” property and thus contribute to their intrinsic capacity for multidrug resistance. Indeed, SP cells have been identified in different subtypes of human breast cancer cells, which display a more tumorigenic phenotype than the cells do not readily efflux vital dyes [52, 58, 59, 60]. These studies further identified an increased expression of ABC transporter genes, especially the breast cancer resistance protein BCRP (also named as ABCG2) and the multidrug resistance protein 1 (also named ABCB1), in the SP cells that containing BCSC activities. The expression of ABC transporter proteins in BCSCs suggests that these cells play important roles in development of drug resistance in breast cancer. Interestingly, the SP phenotype was found to be more prevalent in luminal subtype of breast cancer cells compared to the triple negative (ER−, PR−, and HER2−) subtype, and HER2 expression, which specifically upregulates ALDH+ luminal BCSCs [121], was significantly correlated with the occurrence of SP phenotype [60]. The strong association of SP phenotype with luminal BCSCs may contribute to the increased resistance of chemotherapy found in ALDH1+ breast cancers [119].

5.2.2 ALDH Enzymatic Activity Expressing in BCSCs Directly Involves in Drug Resistance

Aldehyde dehydrogenases (ALDH) are a family of NAD(P)-dependent enzymes involved in detoxifying a wide variety of endogenous and exogenous aldehydes [122]. Mounting evidence suggests that ALDH activity can be utilized either alone or in combination with different cell surface markers to identify CSCs in a wide varieties of tissue malignancies including leukemia [70] and solid tumors, ranging from those of the breast [22], colon [72, 73], bladder [123], prostate [74], lung [71], pancreas [124], head and neck [38], ovary [125], and melanoma [45]. Recent studies have indicated that ALDH enzymes can directly regulate resistance of cancer cells to many cytotoxic drugs. The role of ALDH in drug resistance was first observed for the alkylating agent, cyclophosphamide. Overexpression of the ALDHA1 gene resulted in a significant increases in cyclophosphamide resistance in transduced L1210 and U937 cells [126] and K562 leukemic cell lines [127]. Conversely, downregulation of ALDH1A1 by antisense RNA resulted in increased sensitivity of tumor cells to 4-hydroperoxycyclophosphamide (4-HC), an active derivative of cyclophosphamide [128]. Overexpression of ALDH1 has been observed in many drug resistant cancer cell lines as well as in patients with drug resistant cancers including breast cancer [119, 129]. In a retrospective study, ALDHA1 activity was significantly higher in breast cancer metastatic cells, which developed resistance to cyclophosphamide, compared with sensitive cells [130]. Most recently, Croker and Allan showed that ALDHhiCD44+ BCSCs isolated from MDA-MB-231 and MDA-MB-468 cell lines are more resistant to chemotherapy (doxorubicin/paclitaxel) or radiotherapy than ALDHloCD44 cells, and that inhibiting ALDH activity through all-trans retinoic acid (ATRA) or the specific ALDH inhibitor diethylaminobenzaldehyde (DEAB) sensitizes ALDHhiCD44+ BCSCs to treatment [131]. Together, these studies strongly suggests that ALDH activity expressed in BCSCs plays an important role in mediating BCSC resistance to conventional cancer therapy, and inhibitors of ALDH enzymes may serve as potential therapeutic agents to prevent treatment resistance in breast cancer.

5.2.3 BCSCs Exhibit Low Reactive Oxygen Species Levels and Radioresistance

Increased production of reactive oxygen species (ROS) has long been described as a critical mediator of radiation and chemotherapy induced cytotoxicity to cancer cells by inducing DNA damage such as single-stranded and double-stranded breaks [132, 133, 134]. In neural stem cells (NSCs) and hematopoietic stem cells (HSCs), protection from oxidative stress is critical for the maintenance of their self-renewal [135, 136, 137], and mice deficient in ATM kinase or FoxO1, FoxO3, and FoxO4 transcription factors exhibit elevated ROS levels in the HSC compartment which result in a rapid extinction of HSCs [136, 138, 139]. Similar to the paradigm found in NSCs and HSCs, a recent study by Diehn et al. convincingly demonstrated that human and mouse BCSCs, similar to their normal tissue counterparts, have increased expression of free radical scavenging systems and maintain low levels of ROS that result in less DNA damage and radioprotection [114], providing a possible explanation for tumor recurrence after radiation therapy. These findings also suggest that the self-renewal potential of CSCs in different tumor tissues may be exquisitely sensitive to ROS levels, and this liability might represent a CSC “Achilles’ heel” for future therapeutic exploitation to eliminate these lethal seeds of cancer. Indeed, pharmacological depletion of ROS scavengers in BCSCs markedly decreases their clonogenicity and results in radiosensitization [114]. Accordingly, treatment of human acute myeloid leukemia (AML) with parthenolide, a naturally occurring compound that induces ROS, effectively targets AML stem and progenitor cells for apoptosis [140].

5.2.4 BCSCs Have Enhanced DNA-Damage Repair Responses

A number of studies have suggested that resistance to chemo or radiation therapy observed in CSCs or tumor-initiating cells is mediated by altered DNA damages responses. The Ataxia Telangiectasia Mutated (ATM) gene maintains genomic stability by activating a key cell-cycle checkpoint response to DNA damage, telomeric instability or oxidative stress. In glioblastoma following radiation, increased activation of ATM kinase pathway has been reported in CD133+ CSCs, contributing to tumor radioresistance through preferential activation of the DNA damage checkpoint response and an increase in DNA repair capacity [141]. Blocking the checkpoint kinases Chk1 and Chk2 led to enhanced killing of CSCs by ionizing radiation [141], suggesting that pharmacological approaches targeting DNA damage checkpoint kinases may overcome this intrinsic resistance. The enhanced ATM signaling activity in CD24CD44+ BCSCs compared to non CD24CD44+ tumor cells following radiation was also found in breast cancer cell lines and primary human breast cancer cells [142], and treatment with an ATM inhibitor effectively decreased the radiation resistance of CD24CD44+ BCSCs [142], suggesting that targeting ATM signaling may provide a new tool to abolish the radiation resistance of BCSCs. In mouse models of breast cancer, BCSCs deficient in p53 showed accelerated DNA repair activity, as well as high levels of Akt and Wnt signaling [143] and pharmacological inhibition of the Akt pathway by perifosine, an Akt inhibitor, reduced the content of BCSCs in the p53 null tumor and sensitized BCSCs to radiation treatment [143].

6 BCSCs and Tumor Metastasis

In human breast cancer tissues, one of the first identified markers to enrich BCSCs was CD44, a transmembrane glycoprotein serving as a major adhesion molecule and receptor for the extracellular matrix especially extracellular glycosaminoglycan hyaluronic acid [144]. CD44 and its alternative splicing variants have been shown to form co-receptor complexes with various receptor tyrosine kinases to modulate diverse cellular signaling events and regulate cell proliferation, migration, and invasion [144, 145, 146, 147, 148]. In line with these studies, expression of CD44 in breast cancer cells has been demonstrated to potentiate the adherence of cancer cells to bone marrow endothelial cells [149] and promotes bone metastasis by enhancing tumorigenicity, cell motility, and hyaluronan production [150]. Besides a direct role in promoting tumor cell motility, CD44 has been implicated to play a role in EMT, a process critical for embryogenesis but abnormally activated during cancer metastasis and recurrence. In a recent study, Brown et al. demonstrated that a shift in CD44 expression from variant isoforms (CD44v) to the standard isoform (CD44s) was essential for mammary epithelial cells to undergo EMT, and was required for the formation of breast tumors that display EMT characteristics in mice [151]. The role of CD44 in promoting EMT and mammary tumorigenesis is in agreement with another report showing that EMT generates cells with stem cell properties [61]. The nature of CD44+ BCSCs as the cells undergo EMT strongly suggests these cells are likely to play critical role in tumor metastasis.

The increased expression of the cell adhesion/invasion molecule CD44 in BCSCs suggests that these cells not only play important roles in tumor initiation (by definition, BCSCs are highly tumorigenic cells when implanted into immunodeficient mice), but also act as major players for tumor progression and metastasis. Indeed, several other cell adhesion molecules including CD29 (β1 integrin), CD49f (α6 integrin), and CD61 (β3 integrin) are later identified to enrich both normal and malignant mammary stem/progenitor cells [8, 9] in human and mouse mammary epithelium. In mouse breast cancer model derived by MMTV-PyMT oncogene, the subset of tumor cells defined by CD24+CD29hiCD61+ is not only highly tumorigenic but also has significantly higher motility as shown in Boyden chamber assay [152].

In contrast to elevated expression of adhesion molecules including CD44, CD29, and CD49f in BCSCs, another marker of BCSCs is the negative or low expression of heat stable antigen, CD24 [21]. Interestingly, high levels of CD24 expression have been shown to suppress CXCR4 activity [153], whereas the CXCL12/CXCR4 signaling axis has been implicated in trafficking of breast cancer cells to sites of metastasis [154]. The negative or low expression of CD24 in BCSCs may enable these cells to have high CXCL12/CXCR4 signaling activity and facilitate their metastatic spreading and ability to colonize distant organs.

In addition to the roles of BCSC markers in regulation of cancer cell migration, invasion and metastasis, the CD24CD44+ BCSCs isolated from a variety of breast cancer lines have been shown to express higher levels of inflammatory cytokines and proteins associated with invasion and bone metastasis, including IL-1α, IL-6, IL-8, and urokinase plasminogen activator [155]. Recent gene profiling studies also suggested that BCSCs possess an “invasiveness” gene signature that is associated with decreased metastasis-free survival [156]. In support of these findings, Balic et al. have demonstrated that micrometastasis isolated from the bone marrow of early-stage breast cancer patients are highly enriched for the cells displaying the CD24CD44+ BCSC phenotype [157]. Using human-in-mouse breast cancer orthotopic models combined with noninvasive imaging techniques, Liu et al. further showed that CD24CD44+ BCSCs isolated from human breast tumors are involved in spontaneous metastasis [158]. Recent studies of CTCs in breast cancer blood samples also revealed that these CTCs share a BCSC phenotype [159].

Although BCSCs were initially identified with an EpCAM+CD24CD44+ phenotype [21], subsequently it was determined that human normal and malignant mammary epithelial cells with increased ALDH activity have mammary stem/progenitor properties [22]. In breast carcinomas, high ALDH activity identifies the tumorigenic cell fraction, capable of self-renewal and of generating tumors which recapitulate the heterogeneity of the parental tumor [22]. ALDH1 expression represents the first independent prognostic marker to predict metastasis and poor patient outcome in inflammatory breast cancer (IBC), an angio-invasive form of breast cancer associated with a high incidence of early nodal and systemic metastasis [160]. ALDH+ cells as assessed by Aldefluor assay from various mammary carcinoma cell lines also display increased invasive characteristics as well as an increased ability to metastasize following intra-cardiac injection in NOD/SCID mice [161].

The properties of two BCSC states, including their invasive capacity assessed by Matrigel invasion assay, have recently been characterized [23]. These studies demonstrate that MET-like ALDH+ BCSCs are proliferative, while EMT-like CD24CD44+ BCSCs are more dormant in cell cycling. Within the two BCSC populations, CD24CD44+ cells were significantly more invasive than ALDH+ cells and cells that displayed all three stem cell markers have the greatest invasive capacity. These findings support the proposed model (Fig. 2b) in which EMT BCSCs that are found at the tumor invasive front enter the blood circulation where ultimately they may metastasize to distant sites. These micro-metastases are non-proliferative and remain dormant until they are induced into a transitional state to convert back to an MET “self-renewing” phenotype which can generates additional BCSCs as well as the more differentiated cells which form the tumor bulk. This model is further supported by studies which have demonstrated that both CTCs as well as disseminated tumor cells (DTCs) in the bone marrow of breast cancer patients show significant increase in the CD44+/CD24−/low cell population that are non-proliferative [155, 157]. One of the characteristics of EMT BCSCs is very low or absent expression of the epithelial marker EpCAM. Therefore, using anti-EpCAM antibodies to capture CTCs in methods such as the CellSearch™ assay may miss important populations of CTCs which display an EMT phenotype. The importance and implications of CTCs as well as alternative CTC capture technologies are discussed below.

7 Circulating Tumor Cells (CTCs)

7.1 Clinical Importance of Circulating BCSCs

More than 100 years ago, CTCs were first described by Thomas Ashworth, an Australian physician. In 1889, Steve Paget proposed “the seed and soil hypothesis” that explains the existence of cross talk between the seed (cancer cells) and the soil (microenvironment of distant organ) during the process of metastasis [162]. Prior to distant organ metastasis, cancer cells must route from primary tumor tissue to the blood stream. These so called CTCs have been documented to be enriched in CSCs [163, 164, 165]. Fortunately, most of cancer cells that enter the bloodstream from a primary tumor die as a consequence of shear force or anoikis, or are eliminated by the immune system. Therefore, only a small number of CTCs survive to extravasate to a distant organ. Successfully disseminated cancer cells may either grow to form metastatic colonies or remain dormant for many years [164].

Initially most studies focused on identifying CTCs of an epithelial phenotype [166, 167], but subsequent studies revealed that CTCs consist of several subpopulations including CSCs with different characteristics. Overall, the concept of a noninvasive liquid biopsy to explore mechanisms involved in metastasis makes CTCs an active area of cancer research. In xenograft models, metastasis initiating cells (MICs) within the CTCs have been shown to be CD44 positive. These CD44+ MICs also have been demonstrated to express other markers such as MET and CD47 [158]. In a cohort study, it has been reported that the number of CTCs with an EpCAM+/CD44+/MET+/CD47+ expression profile increased with the clinical progression without significant alterations in the bulk CTC numbers [168]. These findings that CD47 and MET are expressed in both MIC-containing CTCs and metastatic disease may have implications for developing new therapies. In another animal study, it was demonstrated that EpCAM/ALDH1+/CD45 CTCs isolated from blood sample of breast cancer patients were capable of metastasizing to brain and lung tissues of recipient mice [169]. These CTCs showed a HER2+/EGFR+/HPSE+/Notch1+ gene expression profile that are enriched in brain metastasis. They further showed that these EpCAM-negative breast cancer CTCs with overexpression profile of brain metastasis markers increased brain metastatic capabilities and frequency. These data on metastatic potential of EpCAM-negative CTCs may explain the reason for undetectable CTCs by the CellSearch™ (Veridex LLC) method in a majority of breast cancer patients who developed brain metastasis [169].

Recent studies suggest that the CTCs and primary tumors may display different phenotypes which may reflect tumor evolution as well as differential expression of markers. In a serial CTCs monitoring study in human breast cancer, mesenchymal cells were found to be highly enriched in patients’ CTCs. These isolated CTCs demonstrated expression of TGFβ pathway components and the FOXC1 transcription factor [170]. Interestingly, the same study showed that changes in the CTCs content and type in serial blood samples was correlated with primary response to therapy followed by resistance and finally disease progression.

It has been reported that HER2 plays a prominent regulatory role in the breast CSCs [118]. Moreover, in luminal breast cancers, HER2 may be selectively expressed in CSCs in the absence of HER2 gene amplification [121]. These findings might account for the reports of detection of HER2 expressing CTCs isolated from blood sample of women with HER2 “negative” breast cancers [171]. This can also explain the surprising observation that the clinical benefit of HER2 blockade in the adjuvant setting may be extended to woman whose breast tumors do not display HER2 gene amplification. In an artificial neural network (ANN) analysis of CTCs from metastatic breast cancer (MBC) patients, a linear increase of risk of death was reported with increasing CTC counts across a panel of molecularly different tumor subtypes [172]. However, the CTCs prognostic effect was less evident in the HER2+ patients following targeted therapy [172]. These findings support the concept that the number of CTCs along with the biologic characteristics may have prognostic significance.

7.2 Isolation Methods for Characterization of CTCs

A wide variety of methods has been developed to isolate and characterize CTCs. Most of these techniques are either based on immunodetection of surface markers on CTCs or size and morphology of cancer cells. Of these approaches, immunoaffinity based capture devices are more widely used for CTC isolation. In these methods, distinct antibodies are used against surface markers that are expressed exclusively on tumor cells but not present on the blood cells. Anti-EpCAM antibody is the most commonly used antibody for isolation of CTCs. In these techniques, the antibody may be chemically tethered to magnetic beads or a capture surface, ultimately allowing isolation of CTCs. These immunoaffinity based techniques which are highly specific for CTCs isolation have demonstrated the prognostic, monitoring, and molecular diagnostic potentials of CTCs. Immunocapture systems have evolved from the macroscale operation to a host of microfluidic devices being developed today.

A successful example of macroscale immunocapture system is the CellSearch™ (Veridex LLC), which is FDA approved for use with metastatic breast [164], colon [173], and prostate cancer patients [174]. The antibody in this system is raised against an epithelial surface marker, EpCAM, allowing for the separation of epithelial cancer cells from the blood. A study on blood samples from metastatic patients, patients with nonmalignant disease and healthy volunteers showed the ability of CellSearch to detect as few as 2 CTCs/7.5 ml blood in 36 % of metastatic patients versus only 0.3 % of healthy and nonmalignant samples [175]. Despite being sufficiently robust for FDA approval, the CellSearch™ system lacks high purity, recovery and sensitivity that new and improved cell capture systems need to be developed. Considering the advantages of minimized footprint, cost, and reagent expenditures coupled with a library of well-documented fabrication methods [176], it is not surprising that the field of CTC research turned to microfluidics. Microfluidic devices have been used for biological analysis in the form of “labs-on-a-chip” for the polymerase chain reaction (PCR), molecular separation, and immunosensing [177, 178] through the construction of systems with small scale mixers, pumps, reservoirs, and valves [179].

Microfluidics first appeared in field of CTC isolation in 2007 with the introduction of the CTC-chip [167]. The main advantages of this device include isolation of viable cells for further potential downstream analysis, and the ability to use the whole blood without further processing. This device has been used in the clinical setting by processing 116 samples with CTCs detected in 99 % (115 of 116) of patient samples. The immunocapture-based microfluidic CTC separation devices have been further engineered with respect to capture antibodies, geometries, and materials. The circular posts in the CTC-chip have been developed in the geometrically enhanced differential immunocapture (GEDI) chip [180]. In addition, the use of the transparent and inexpensive PDMS bonded to a transparent glass slide facilitates imaging that allows on-chip FISH (fluorescence in situ hybridization). Polymers that are cheap and easy to pattern and mold can also be used as a capture substrate in microfluidic devices such as high-throughput microsampling unit (HTMSU) [181] and the “Herringbone-chip” [182]. More recently, a novel nanomaterial graphene oxide (GO) patterned onto a silicon capture surface and conjugated with anti-EpCAM antibody that yielded both high CTC capture efficiency and unprecedented purity has been developed [183].

Although immunoaffinity techniques have the advantage of high specificity, they are limited to only isolate cells with the specified antigens. Therefore, these methods may not be appropriate for isolating all different types of CSCs at the same time from whole blood due to lack of unique surface marker that can distinguish CSCs from hematopoietic cells. Considering these limitations, researchers have focused on development of “label–free” isolation methods based on the biophysical properties of the CTCs including membrane potential, dielectric properties, size of cancer cells relative to blood cells, and their difference in adhesion preferences.

In order to move away from immunoaffinity based isolation methods, researchers focused on size based technologies to isolate CTCs derived from solid tumors that are large enough to be distinguished from blood cells. Size based filtration methods have emerged [184, 185] in which the cells are passed through pores etched in membranes, physically retaining larger cells on the top of the membrane. In this respect, an efficient membrane microfilter device was made of parylene-C for the isolation of prostate cancer cells from whole blood [186]. Moreover, researchers used two-layer membranes to filter viable prostate and breast cancer cells [187]. Two problems associated with size based methods when increasing volumes of blood are (1) easy clogging and (2) the requirement for diluting whole blood before filtering. To address these issues, a microfluidic filtration system consisting of a serpentine channel interconnected with two outer filtrate channels with rectangular apertures was developed which could isolate breast cancer cells spiked into whole blood with 50–90 % recovery rate [188]. Despite improved processing speed as compared to immunoaffinity capture methods, the volume of blood sample that can be processed without sacrificing efficiency and purity is still limited. Different approaches have been used to improve high throughput analysis of blood samples that are based on shape, geometry and distribution of microchannels [189, 190, 191]. Inertial sorting has been coupled with both positive and negative cell sorting in the CTC-iChip [192]. In this case, as with the negative sorting approach that does not require labeling of the CTCs, researchers can isolate cells from multiple cancers and phenotypes. Although these inertial microfluidics based techniques offer high throughput and sensitivity, they suffer from a lack of sufficient specificity of isolated cells. An array of precisely spaced posts has been introduced to capture cells that are not flexible enough to travel through the gaps. These cancer cells were shown to have CD44+/CD24/claudinlow gene expression pattern and tumor initiating capacity [193]. Because of the flexibility of blood cells, this approach can be used as a secondary enrichment step following CTC separation from the blood using other label-free methods. Despite these limitations in microfluidic devices, this class of techniques offers the ideal platform for the label-free isolation of CSCs within the CTCs.

Future methods for CTC capture will aim to solve these problems, ideally increasing the specificity, yield, and throughput. These approaches will in all likelihood take the direction of integrated modules that allow the advantages of multiple techniques and the use of nanomaterials. Integrated devices such as the CTC-iChip use both inertial and immunomagnetic sorting, making feasible the selection of cells of multiple tissue types, stages, and transitional phenotypes. These devices can integrate preprocessing into the device and increase throughput while simultaneously increasing the populations of cells targeted.

The ultimate goal of CTC isolation and characterization is to achieve maximal clinical utility. Therefore, it is essential for the isolation technologies to provide consistent results optimized for yield and purity. More importantly, the isolated CTCs should represent the heterogeneity of the primary tumor cells as well as the alterations necessary to allow the disseminated cells to survive within the blood circulation and finally form metastases at secondary sites. The sensitivity and specificity of isolation methods are key factors to isolate CTCs that are free from blood cells to allow useful downstream experiments. Furthermore, the quantity of isolated CTCs is important for functional analysis and meaningful biological assays. Currently the clinical utility of CTCs is hindered by the low number of CTCs, and hence, the ability to process larger volume of blood sample can change this paradigm. To achieve these goals, isolation methods should be rapid and be able to provide reliable data with respect to the number and cellular phenotypes of CTCs to predict the disease status and treatment response.

8 Patient Derived Xenograft Models for Cancer Stem Cell Research

Despite marked progress in our understanding of cancer biology, the translation of research findings into new therapies for cancer remains an enormous barrier with a failure rate of about 90 % for new oncology drugs in the clinical settings. Part of this high failure rate is due to the lack of in vivo preclinical models that authentically reflect patient tumor biology. Historically, in vivo experimental therapeutic research has been largely dependent on either genetically engineered mouse models, or “xenograft” transplantation models, wherein established human cancer cell lines are transplanted into immune-compromised host mice. These models, although useful and informative, only partially recapitulates the primary tumors in patients. To overcome these limitations, use of patient derived xenograft (PDX) models through engraftment of actual tumor tissues into immune-deficient mice is increasing in cancer biology and preclinical drug testing. These tumor xenografts generally retain the morphology, cellular heterogeneity, genetic background and molecular expression profiles of the original patient tumors. In addition, serial passage and expansion of patient tumors through immune-deficient murine hosts permits ongoing propagation of tumor lines and the study of tumor biology without subjecting PDX tumor cells to the stressful and compromising culture conditions encountered in vitro.

Early attempts to establish human primary breast cancer tissue xenografts used athymic (nude) mice that lack T cell immunity or NOD/SCID mice that lack both T- and B-cell function but retain innate cellular immunity. However, a technical hurdle frequently presented in these studies is the low engraftment rate and the difficulty to maintain the transplantable tumor xenografts over time. The use of “humanized” NOD/SCID mice by introducing an immortalized human fibroblast cell line into the mammary fat pad before transplantation has greatly increased the efficiency to establish PDX models of human breast cancer [194]. Using primary tumor cells transplanted orthotopically in humanized cleared fat-pad of NOD/SCID mice, many serially transplantable tumor xenotransplants have been established. For example, our group has established several widely used xenotransplants derived from independent human breast cancers including MC1, UM1, UM2, and UM3. The ERPRErbB2 MC1 and ER+PR+ErbB2 UM2 xenotransplants were derived from metastatic tumors of pleural effusion and ovarian metastasis respectively, while the UM1 and UM3 were from ERPRErbB2 primary tumors [22]. Taking advantage of these PDX models, Ginestier et al. (2007) showed that the Aldefluor-positive tumor cell population isolated from human breast tumors xenografts has CSC properties [22].

The development and use of PDX models represents the gold standard to assess tumor stem cell activities. Although BCSCs can be propagated in vitro through non-adherent culture conditions, these culture conditions cannot recapitulate the tumor microenvironment in vivo. Tumor xenograft assays in mice measure not only tumorigenic activity but also the self-renewal and maintenance of tumor stem cell component presented in the original tumor. In the landmark study, Al Hajj et al. (2007) transplanted a number of patient tumors into the thoracic mammary fat-pad of etoposide- and estrogen-pretreated NOD/SCID mice and successfully identified a tumorigenic subpopulation in breast cancers with stem cell activities [21]. More recently, Liu et al. established additional models from primary and metastatic tumors, with the majority of these PDXs generating lung micrometastases [158].

In contrast to the use of immortalized human fibroblast cell line to “humanized” NOD/SCID mice, Derose et al. (2011) have showed that implantation of human mesenchymal stem cells (MSCs) with tumor transplants into cleared fat pad of NOD/SCID mice increased tumor growth and stability of the tumor grafts by promoting angiogenesis [195]. These studies of PDX models are in line with the studies of Liu et al. (2011) who demonstrated that ALDH+ stem cells in patient breast tumors are regulated by MSCs, and co-transplantation of MSCs via intratibial injection accelerated tumor growth by increasing the breast CSC population [161]. Besides the frequently utilized NOD-SCID mice, use of SCID/Beige and NSG mice that lack T-cell, B-cell, and NK cell function for xenotransplantation has been proved to increase the efficiency to establish tumor xenografts of human breast cancers. Recently, Michael Lewis’ team at Baylor College in Houston established a large cohort of 32 stably transplantable xenograft lines in SCID/Beige and NSG mice. These tumor xenograft lines represent different breast cancer subtypes including triple negative (ERPRHER2), HER2-positive (ERPRHER2+), ER-positive (ER+PRHER2), ER/PR double positive (ER+PR+HER2), and triple-positive (ER+PR+HER2+), and provide a renewable, quality-controlled tissue resource for preclinical studies to investigate treatment response and metastasis [196]. In another study, Charafe-Jauffret et al. (2013) have also established primary breast tumor-derived xenografts in NSG mice that encompass the main diversity of human breast cancers and retain the major clinicopathologic features of primary tumors [48]. In this study, 20 PDXs were established from 74 primary breast tumors with an engraftment rate of 27 %, and 13 PDX lines were successfully maintained through serial passages with a transplantable rate of 17.5 %. Interestingly, successful engraftment of patient tumor was found to correlate with the presence of ALDH+ BCSCs, which predicted poor prognosis in patients. The correlation of ALDH+ BCSC content with increased tumor engraftment rate is in agreement with a number of other studies showing that high grade TNBCs and metastatic tumors (which contain higher BCSC content) have higher graft take rates than low grade ER+ luminal tumors [196, 197, 198, 199, 200]. These studies have further strengthened a critical role of BCSCs in initiating and maintaining tumor growth.

Endocrine therapy has proven very successful and yet only 50 % of all estrogen receptor-positive breast cancers respond to endocrine therapies [201]. Recent development of PDX models of hormone-driven breast cancer has provided a valuable assay to study hormone dependence and resistance of luminal breast cancers. Historically, tumor xenograft strategies for hormone-driven luminal breast cancers have generated limited success. This is largely due to fact that luminal tumors inherently have lower pathological grades and slower growth rates. Recent efforts focusing on generation of ER-positive luminal tumors has increased the number of stable luminal PDXs. DeRose and colleagues established four luminal B PDXs with co-transplantation of human MSCs [195]. Cottu et al. (2012) established a panel of eight luminal breast carcinoma xenografts by engrafting 423 tumors including 314 ER+ tumors, with a tumor take rate of 2.5 %, which is much lower than for non-luminal tumors (2.5 vs. 24.7 %, P < 0.0001). Histological and immunohistochemical studies on patient’s tumors and xenografts revealed striking similarities in the tumor morphology as well as in the expression level of ER, PR, and HER2. In addition, six luminal models had different sensitivities to hormone therapy, thus exhibiting heterogeneity similar to what is observed in the clinic [202]. Kabos et al. (2012) recently described the development and characterization of five transplantable luminal ER+ breast cancer xenografts, derived from both primary untreated tumors and late stage metastases. Four of the tumor xenografts were found to be estrogen-dependent and tamoxifen or estrogen withdrawal abrogated estrogen-dependent growth [203]. Interestingly, examination of the BCSC population in the five ER+ PDX models revealed that tumor xenografts from untreated primary tumors generally contained low CD24−/lowCD44+ BCSCs. In contrast, xenograft derived from late stage metastases contained a significant higher CD24−/lowCD44+ population [203]. These studies further evidenced an important role of BCSCs in promoting tumor metastasis in luminal tumors.

PDX models of human breast cancer are increasingly used in preclinical study to evaluate the impact of potential breast cancer therapies to eliminate BCSCs. Ginestier et al. (2010) demonstrated that reparixin, a small-molecule inhibitor of the IL8 receptor CXCR1, has the potential to selectively target BCSCs. Administration of reparixin together with Docetaxel, but not Docetaxel alone, selectively decreased BCSC population in tumor xenografts from three different patients (MC1, UM2, and UM3) and reduced tumor growth and metastasis [204]. The signal transducer and activator of transcription 3 (Stat3) serves as an oncogene activated in many cancers including breast, prostate, lung, head and neck and colon, liver, pancreas, and multiple myeloma [205, 206, 207]. Dave et al. (2012) recently showed that targeting the SH2 domain of Stat3 with a novel small molecule inhibitor decreased the percentage of cells expressing BCSC markers (CD24−/lowCD44+ and ALDH+) and mammosphere formation in phospho-Stat3 overexpressing human breast cancer xenografts in SCID-beige mice [208]. Moreover, treatment of Stat3 inhibitor plus docetaxel resulted in a fourfold improvement in recurrence-free survival relative to docetaxel alone in the chemo-resistant tumor model [208]. These findings provide a strong impetus for the development of selective Stat3 inhibitors to target BCSCs in phospho-Stat3 overexpressing breast cancers.

The delta-like 4 ligand (DLL4) is an important component of the Notch pathway and contributes to stem cell self-renewal and vascular development. Previous studies have indicated that inhibition of DLL4 resulted in broad spectrum antitumor activity in cancer cell line-based xenograft models [209, 210, 211]. Using specific antibodies targeting DLL4, Hoey et al. (2009) demonstrated that specifically inhibiting human DLL4 in the tumor xenograft models, either alone or in combination with the chemotherapeutic agent irinotecan, reduced CSC frequency, as shown by flow cytometric and in vivo tumorigenicity studies [212]. In another study, treatment with a gamma secretase inhibitor, GSI, has been shown to reduce BCSCs in MC1 and other tumorgrafts by inhibition of the Notch pathway [213]. In addition, treatment with GSI enhanced efficacy of docetaxel and reduced BCSCs [213]. Using PDX model of breast cancer, effect of erythropoietin (EPO) on sensitivity of chemotherapy on BCSCs was also evaluated. EPO administration counteracted the effects of chemotherapeutic agents on BCSC-derived orthotopic tumor xenografts and promoted metastatic progression both in the presence and in the absence of chemotherapy treatment. This suggested that EPO acts directly on BCSC by activating specific survival pathways, resulting in BCSC protection from chemotherapy and enhanced tumor progression [214]. Taken together these studies suggest that PDX models of human breast cancers, which recapitulate the complexity and heterogeneity of patient tumors, have the advantages that more accurately reflect human breast cancer biology than other existing models. These PDX models also have great potential to facilitate the next phase of drug discovery designed to target BCSCs, and ultimately cure this disease.

9 Key Signaling Pathways Regulating the Self-Renewal of CSCs

Several signal transduction pathways such as Wnt, Notch, and Hedgehog and molecules such as Bmi-1 are known to regulate self-renewal pathways in normal stem cells, while in CSCs these pathways are normally dysregulated due to accumulated mutations and epigenetic changes. Understanding the signaling pathways through which CSCs regulate their self-renewal and maintenance and hence tumor growth and metastasis is important for developing targeted therapies to abrogate CSCs. Conventional cancer therapies normally target aberrant pathways in the rapidly proliferating bulk tumor cells, but often spare the CSCs leading to tumor recurrence and metastasis. Therefore, the design of new therapies must be based on targeting the signaling pathways that affect both CSCs as well as bulk tumor cells. Here, we discuss the main pathways that are involved in CSC self-renewal along with their potential therapeutic implications.

9.1 Hedgehog Signaling Pathway

Hedgehog (Hh) is a family of secreted ligands including Sonic, Indian, and Desert which activate Gli transcription factors through interaction with HIP1, Ptch-1, and Ptch-2. Hh signaling regulates body pattern formation, cell proliferation, cell fate determination, and stem/progenitor cell maintenance [215]. Increased expression of Bmi-1 as well as the Hh pathway components Ptch1, Gli1, and Gli2 has been shown in CD24CD44+ BCSCs compared to non-stem cells [216]. Moreover, overexpression of the Hh target Gli2 in human mammary stem/progenitor cells enriched in mammosphere culture produces ductal hyperplasias when these cells are implanted into the humanized fatpads of NOD-SCID mice [216]. These studies indicate that the hedgehog pathway and Bmi-1 play important roles in regulating self-renewal of normal and tumorigenic MaSC/progenitor cells.

Initially, the Hh pathway was targeted using cyclopamine, a steroidal alkaloid that downregulates Gli1 by binding to Smo and hence suppressing the growth of breast cancer cells [217]. Subsequently, new Hh inhibitors have been developed by chemically modifying cyclopamine [218]. At present, GDC-0449 (Vismodegib, trade name: Erivedge), the first Hh pathway inhibitor approved by FDA [219], is undergoing clinical trials in combination with the Notch signaling inhibitor RO4929097 (a gamma-secretase inhibitor, GSI) for metastatic breast cancers where tumors cannot be surgically removed ( It would be particularly interesting to see the effect of these Hh inhibitors on CSCs as the Hh pathway may be activated in CSCs in response to chemotherapy or during recurrence. Since Hh signaling also imparts chemoresistance [220], the most effective cancer therapy would likely include a Hh inhibitor along with cytotoxic chemotherapy.

9.2 Notch Signaling

Notch signaling involves four homologous transmembrane receptors Notch1-4. Upon binding to their cognate ligands (Delta, Delta-like, Jagged1, and Jagged2), the intracellular domain (ICD) of Notch is cleaved which then translocates into the nucleus to activate its target genes. Notch signaling has emerged as a key regulator involving stem cell maintenance, cell-fate specification, and differentiation [221] and dysregulated Notch signaling has been implicated in a number of human malignancies [222, 223]. In a MMTV mouse model, expression of an activated Notch-related int-3 transgene was shown to interfere with cell differentiation and induce neoplastic transformation in the mammary gland [224]. In a recent study, knockdown of the canonical Notch effector Cbf-1 in MaSC-enriched population was found to increase stem cell activity whereas constitutive Notch signaling specifically targeted luminal progenitor cells for expansion, leading to hyperplasia and tumorigenesis [225]. In human breast cancers, coexpression of JAG1 and NOTCH1 is associated with poor overall survival [226]. In ESA+ CD24CD44+ BCSCs, Notch-4 and Notch-1 activity was found to be eight-fold and four-fold higher respectively compared to the differentiated bulk tumor cells [227]. Expectedly, pharmacologic or genetic inhibition of Notch1 or Notch4 reduced stem cell activity in vitro and reduced tumor formation in vivo [227]. Elevated Notch-1 signaling also contributes to drug resistance as downregulation of Notch-1 signaling in human breast cancer cells increases chemosensitivity to doxorubicin and docetaxel [228].

Several important oncogenic pathways such as ErbB2, Jak/Stat, TGF-β, NF-κB, Wnt, and Hedgehog interact with the Notch pathway [229]. For example, ErbB2 has been shown to induce Notch-1 activity through Cyclin D1 induction [230]. Combined treatment of DAPT, a Notch inhibitor with ErbB2 inhibitor Lapatinib effectively targets stem/progenitor cells both in vitro and in vivo in breast ductal carcinoma in situ (DCIS) [231]. Another study showed that Notch-1 signaling is decreased in ErbB-2 overexpressing SKBR3, BT474 and MCF7/HER2 cells and that HER2-targeted therapies using trastuzumb or lapatinib reactivated Notch-1 and rendered them sensitive to GSIs [232]. These studies suggest that combined treatment of GSI with HER2 targeted therapies may be more beneficial and could potentially reverse the resistance of HER2 targeted therapies especially in CSCs. The Wnt pathway also interacts with Notch through Wnt/TCF target Jagged-1, a Notch ligand, and Mel-18, a negative regulator of Bmi-1. Knockdown of Mel-18 has been shown to enhance the self-renewal of BCSCs whereas its overexpression inhibited the number and self-renewal activity of BCSCs. Mel-18 blockade upregulated Jagged-1 expression and consequently activated the Notch pathway [233]. The activation of Notch activity further activates the Hedgehog pathway and increases expression of Ptch and Gli [216]. Together these studies suggest that treatments aimed at molecules that affect multiple stem cell pathways could present a novel strategy for targeted therapies.

9.3 HER2 Signaling

Recent studies have indicated an association between HER2 and ALDH1 BCSC marker expression [118]. In clinical setting, one-third of HER2+ patients do not respond to HER2-targeting agents such as trastuzumab and further resistance could develop after long term treatments. This resistance has been attributed to loss of PTEN, somatic mutations of PI3K, truncation of extracellular domain of HER2, engagement of alternate signaling pathways, alterations in antibody binding to HER2, loss of the apoptotic response, and evasion of the immunomodulatory effects conferred by trastuzumab [234, 235]. Korkaya et al. (2009) mechanistically demonstrated the role of the PI3K/AKT pathway in BCSC regulation whereby AKT activation upregulates the WNT pathway through phosphorylation of GSK-3β and β-catenin on Ser552 leading to localization of β-catenin to the nucleus [236]. Therefore, targeting AKT pathway may be an additional approach to target the BCSCs. In support of this notion, Perifosine, an AKT inhibitor, effectively reduced BCSCs in tumor xenografts [236].

In HER2+ breast cancers, the combination of a PI3K inhibitor with trastuzumab is known to be effective to overcome trastuzumab resistance [237, 238]. Another approach to overcome the resistance to HER2 therapy is to target downstream nodes of PI3K/AKT pathway. It has been shown that blockade of mTOR downstream of HER2 with everolimus (a TORC1 inhibitor) in combination with trastuzumab and a taxane resulted in significant clinical response in metastatic HER2+ breast cancers [239]. Recently, a study by Chakrabarty et al. (2013) demonstrated that combined treatment of the pan-PI3K inhibitor, XL147, and trastuzumab reduced proliferation and triggered apoptosis of trastuzumab resistant breast cancer cells [240]. Further, the CSC fraction within trastuzumab-resistant cells was also reduced both in vivo and in vitro [240], providing the proof of principle that HER2 is involved in stem cell maintenance. Recently, Ithimkain et al. (2013) demonstrated that HER2 is selectively expressed in and regulates self-renewal of CSC population in ER+, HER2 luminal breast cancers [121]. In addition, HER2 expression is increased in luminal tumors grown in mouse bone xenografts, as well as in breast cancer bone metastases as compared with matched primary tumors. This increased expression of HER2 in luminal tumor grown in bone xenografts was mediated by receptor activation of NF-κB (RANK)-ligand in the bone microenvironment [121].

9.4 Wnt Pathway

The Wnt family is a group of secreted glycoproteins that binds to the protein complex of Frizzled and low density lipoprotein receptor related protein 5 or 6 (LRP5 or LRP6). Downstream of LRP5/LRP6 is GSK3β which phosphorylates β-catenin leading to its ubiquitin-mediated degradation. Activation of the Wnt pathway phosphorylates GSK3β and hence stabilizes β-catenin which then translocates to the nucleus activating several oncogenes such as ID2, MMP7, and c-Myc [241]. The noncanonical Wnt pathway is known to act through Rho family small GTPase, calcium and protein kinase A signaling. The Wnt pathway regulates cell fate determination in several tissues including the mammary gland [242]. In LRP5 knockout mammary glands, very few stem or progenitor cells were present compared to wild type mammary glands [243]. Activation of Wnt signaling and its components have been implicated in variety of cancers including breast [244, 245, 246, 247]. A recent study further indicated that increased expression of stemness gene Sox2, by activating the Wnt signaling pathway, promotes resistance in breast cancer cells [248].

Transgenic mice overexpressing Wnt-1 in mammary glands were enriched for epithelial cells expressing progenitor cell markers keratin 6 and Sca1 and tumors that developed in these mice contained cells expressing keratin 6 [249]. This suggests that mammary stem cells and/or progenitors may be the targets for oncogenesis by Wnt pathway. Furthermore, the transforming activity of Wnt effectors was shown to be correlated with their ability to induce accumulation of mammary progenitor cells [250]. The AKT/β-catenin pathway is also activated by antiangiogenic agents such as sunitinib and bevacizumab which drives CSCs expansion through HIF1 alpha [251]. Targeting of the Wnt pathway could be achieved by several approaches. For example, methylation-associated silencing of SFRP1 was shown to inhibit Wnt signaling in breast cancer [252]. In breast cancer cell lines including MCF7, HuL100 and SKBR3, incubation with Wnt1 monoclonal antibody has been used to inhibit Wnt-1 signaling and induce apoptosis [253]. The redundancy between different ligands may suggest that antibody directed Wnt inhibition would not be a successful approach. However, since some cancers have been shown to rely heavily on specific Wnt isoforms, it may be a viable approach in those cancers. For tumors which do not rely on specific Wnt, the use of pan-Wnt inhibitor may be more efficacious. A recent study demonstrated that a soluble ligand binding domain of Fzd8, Fzd8-CRD-Fc, inhibited autocrine Wnt signaling in vitro, as well as in multiple xenograft models [254].

9.5 Regulation of BCSCs by Tumor Microenvironment

The tumor microenvironment involves diverse elements including cancer cells, fibroblasts, endothelial cells, inflammatory cells, and MSCs. Since the tumor microenvironment plays a pivotal role in regulating stem cells; it has been collectively called the stem cell niche [255]. The tumor microenvironment, through paracrine interactions or cytokine networks, may activate the signal transduction pathways described above to regulate CSCs. It has been shown that MSCs recruited from breast stroma or bone marrow [161, 256] interact with BCSCs through cytokine loops involving IL-6 and CXCL7 stimulating the self-renewal of BCSCs [161]. It has been reported that estrogen signaling in the ERα+ non-stem cell compartment stimulates the proliferation of cells within the ERα stem cell compartment [257]. Similarly, in luminal breast cancers, estrogen signaling expands the pool of ERα BCSCs through a paracrine FGF/FGFR/Tbx3 signaling which was inhibited by tamoxifen [258].

Circulating markers of inflammation such as serum C-reactive protein (CRP) and amyloid A (SAA) have been shown to correlate with increased risk of breast cancer recurrence in women after primary therapy [259]. Inflammation is known to be mediated by IL6, IL8, IL-1β, and many other cytokines [260]. Gene profiling studies conducted by our group suggested a role of IL8 in promoting tumor stemness in breast cancer cell lines [47]. Indeed, blockade of IL8 signaling through Reparixin, a small molecular inhibitor of the IL-8 receptor CXCR1, has been shown to selectively target BCSCs both in vitro and in breast cancer xenografts, leading to significantly inhibited tumor growth and metastasis [204]. Reparixin is currently being evaluated in a Phase 2 clinical trial at The University of Michigan. In another recent study, Chakrbarty et al. (2013) showed that the combined treatment of Trastuzumab resistant cells with trastuzumab and PI3K inhibitor, XL147, reduced the CSC stem cell population by inhibiting IL-8 [240].

Besides IL8, several preclinical studies have shown that IL6 expression is associated with tumorigenicity and treatment resistance in breast cancer cells [261, 262]. An epigenetic switch involving NF-κB, Lin28, Let-7 MicroRNA, and an IL6 feedback loop has been reported to link inflammation to mammary epithelial cell transformation [263]. Korkaya et al. (2012) further showed that trastuzumab resistant PTEN knockout breast cancer cell lines had an activated IL6 inflammatory feedback loop leading to expansion of the EMT-like CSCs [264]. This suggests that in addition to the NF-κB, IL6 could also serve as a molecular target to eliminate BCSCs. The Stat3 protein plays a central role in relaying extracellular signals initiated by cytokines and growth factors [205, 265]. Therefore, in addition to IL6 targeting, inhibition of Stat3 could serve as an alternative therapeutic strategy. In ALDH+ and ALDH+/CD44+/CD24 BCSCs, higher level of phosphorylated Stat3 (P-STAT3, Y705) compared to bulk tumor cell populations has been demonstrated [266]. Furthermore, a novel Stat3 inhibitor, LLL12, suppresses ALDH+ and ALDH+/CD44+/CD24 BCSCs in vitro and inhibits tumor growth in mouse xenograft and mammary fat pad models in vivo [266]. Thus, Stat3 may represent a target for therapeutic intervention in BCSCs and inhibition of constitutive Stat3 signaling may provide a novel therapeutic approach.

9.6 Regulation of BCSCs by MicroRNAs

Recently, MicroRNAs (miRNAs or Mirs) have emerged as master regulators of stem cell pathways. Mir93 expression has been associated with downregulating TGFβ signaling along with multiple stem cell regulatory genes including JAK-1, STAT3, AKT3, SOX4, EZH1, and HMGA2 and thus inhibiting stem cell populations [83]. Yu et al. (2007) cell demonstrated that breast tumor initiating cells (TICs) and differentiated cells have low and high levels of let-7 expression respectively [115]. Let-7 overexpression in TICs reduced mammosphere formation in vitro and tumor formation and metastasis in NOD/SCID mice [115]. The same group further showed that miR-30 decreases BCSCs through ubiquitin-conjugating enzyme 9 (UBC9) and integrin β3 [267]. The genomic analysis between BCSCs and bulk cancer cells identified 37 Mirs differentially expressed with miR200c-141, miR-200b-200a-429, and Mir 183-96-182 downregulated in human BCSCs, normal human and murine stem cells. MiR200c expression decreased clonal expansion of breast cancer cells by targeting BMI both in vitro and in vivo [268]. Thus, targeting of Mirs could provide a better approach for eliminating BCSCs since a single Mir could affect multiple pathways.

10 Therapy Implications of CSC-Targeted Approaches

10.1 Cancer Treatment and Prognosis: Why a New Paradigm Is Necessary

Cancer is a leading cause of disease-related mortality worldwide [269]. According to the World Health Organization there were more than 1.6 million new cases of breast cancer in 2010, a dramatic increase as compared to 640,000 cases reported in 1980 [270]. Better early detection methods are unlikely to fully account for such a dramatic rise in incidence in three decades. Over this timeframe progress in the treatment of advanced forms of the most common cancers including those of the breast have shown modest gains in terms of survival metrics. In stark contrast, spectacular advancements have been achieved in the fields of genetics and stem cell biology to characterize and understand the molecular mechanisms that foster cancer. These innovations hold the potential to transform cancer treatment only as long as this growing detailed understanding of the molecular basis of cancer can be translated into clinical benefits. Molecular analyses of tumors reveal a far greater genetic complexity leading to more variation in the pathways that drive tumor growth and metastasis than heretofore recognized. The discovery of genes directly linked to cancer and the molecular pathways influenced by these genes has allowed scientists to draw more precise maps of cancer progression and tautologies. Future drug development efforts will need to put to use this novel understanding to focus targeted therapies that tackle cancer-specific events with unparalleled precision. Currently, the overall 5-year survival rate for all cancers, relative to the anticipated survival from a comparable cohort of people without cancer is 65 % [271]. Statistics like these are sobering, albeit survival and recurrence rates following diagnosis vary greatly and mostly as a function of cancer type and stage at detection. For example, the relative survival rate 5 years following diagnosis of melanoma is greater than 90 % whereas that of cancers of the brain and nervous system is about 35 % [272]. Once a cancer has metastasized, the survival rate declines dramatically. When melanoma is diagnosed at the localized stage, 99 % of patients will survive beyond 5 years. However survival rates drop to 65 % for those diagnosed with regionally metastasized melanoma and to 15 % for those whose melanoma has spread to distant sites [272]. Once a cancer has been diagnosed, conventional treatments vary according to cancer type and severity. Resection, radiotherapy, and systemic treatments such as chemotherapy or hormone therapy represent traditional approaches designed to remove or eradicate rapidly dividing cancer cells [97, 273, 274]. These methods all have limitations for clinical use. Cancer surgeons may not be able to fully remove all of the tumor tissue due to its location, extent of spreading or degree of intercalation with healthy tissue. Radiation and chemotherapy are inherently nonspecific cytotoxic strategies that result in extensive collateral damage to healthy tissue despite their targeting rapidly dividing cancer cells. Recently, several new immune modulating agents that target specific proteins implicated in cancer molecular pathways have been developed for clinical use. These include trastuzumab, a HER2-targeting monoclonal antibody in breast cancer [275], bevacizumab which targets VEGF in colorectal and lung cancer [276], erlotnib and gefitinib which target epidermal growth factor receptor (EGFR) in lung cancer [277], imatinib which targets BCR-ABL in CML [278], and panitumumab and cetuximab which target EGFR in colorectal cancer [276]. These agents demonstrate that a targeted approach is feasible, but they are most often effective only in patients who display specific subclasses of the respective cancers, hence justifying the efforts to develop more personalized medicine. Furthermore, patients can become resistant to these therapies and most systemic immune treatments are only moderately successful unless the cancer is localized. With few exceptions, these agents fail to provide significant clinical benefit at the advanced metastatic stage of the disease. Therefore, a new paradigm is needed for the treatment of resistant and advanced disease. CSC-directed therapies offer the possibility for disease remission without recurrence. In the final section we discuss implications of the CSC hypothesis, supporting evidence for CSC clinical trials and perspectives on how CSCs could impact the development of future cancer therapy.

10.2 Clinical Implications of CSCs

The presence of and the molecular events triggered by CSCs within tumors has important implications for the development of cancer-targeting therapeutics. The CSC hypothesis suggests that tumor regrowth following traditional resection and/or chemo/radio approaches can be arrested if the repopulating cells are destroyed with a selective CSC targeting agent. This is fundamentally important because CSCs are generally found to be more aggressive, invasive and prone to promote metastasis than the bulk tumorigenic cells [83]. Aldefluor+ cells from mammary carcinoma cell lines display increased invasive characteristics as well as an increased ability to metastasize following intra-cardiac injection in NOD/SCID mice [161]. Similarly, Balic et al. observed an increase in the CD24CD44+ breast CSC population in bone marrow metastases in patients with breast carcinomas [157]. As a result, targeting of the CSC clinically could lead to a reduction in metastasis. CSCs are presumed to be responsible for both the radiation- and chemotherapeutic-resistance commonly observed in many cancers [116]. Since stem cells are quiescent and typically cells with slower cycles than differentiated cells, stem cells are intrinsically more resistant to cell cycle altering chemotherapeutic drugs [279]. CSCs also possess a high level of drug transporters and particularly exporters such that the intracellular effective chemotherapuetic concentrations in these cells is lower than in differentiated cells resulting in the apparent decreased overall efficacy of the chemotherapeutic agent administered. Furthermore, the activity of the enzyme ALDH is high in CSCs, and this enzyme has been demonstrated to metabolize certain common chemotherapeutic drugs including cyclophosphamide into forms of metabolites with lower cytotoxicity [69]. To maximize the clinical benefits of drugs that target CSCs, these drugs should thus be used in combination with radiation or standard chemotherapy since CSC-targeting agents are typically not designed to clear the vast majority of differentiated bulk tumor cells and complications from tumor burden alone might adversely complicate patient prognosis. Because the CSC subpopulation of a tumor often constitutes a minor portion of the tumor mass, novel methods need to be implemented such as new trial designs for accurately measuring success and outcome of CSC-targeting agents. Current clinical trials use the so-called RECIST (Response Evaluation Criteria in Solid Tumors) criteria to measure tumor shrinkage. However, tumor regression frequently fails to correlate with improved survival rates [280]. CSC treatment alone would not be expected to elicit major reductions of the parameters that are measured using the RECIST evaluation criteria [19, 280]. CSC clinical trials thus need to be designed to directly measure the effect that the new drug under investigation has on CSCs. This objective could be met by conducting neo-adjuvant CSC trials. For example, CD24CD44+ breast CSCs have been observed to increase after chemotherapy, however combination therapy using BCSC targeting agent such as HER2 and EGFR inhibitor Lapatinib with chemotherapy prevented the increase of BCSCs [117]. A similar outcome could also be accomplished by providing the CSC-targeting agent prior to surgery followed by a CSC biomarker profiling step performed on selected patient biopsies. Practically, cells from these very biopsies could be grown as tumorspheres in order to determine not only a drug’s ability to target CSCs but also which signaling pathways are involved. For effectively destroying CSCs, it is necessary to understand how the various signaling pathways of these cells interact to regulate CSC-specific processes of growth, renewal, differentiation and apoptosis. The Notch, Hedgehog, HER-2/PI3K/Akt/ PTEN, p53 and Wnt signaling pathways have been shown to regulate normal and malignant stem cells [236, 281]. Notably, inhibitors of Notch, Hedgehog, Akt and Wnt targeting agents have been developed by several pharmaceutical companies [19, 282, 283]. Emerging evidence suggests that microRNAs (miRNAs) may play a pivotal role in regulating the genes that control self-renewal, differentiation and division of normal stem and CSC signaling pathways [83].

11 Future Perspectives

The critical roles of CSCs in tumor initiation, progression and recurrence has sparked enormous interest among oncology researchers to explore key signaling pathways that regulate their self-renewal and to develop novel therapeutic strategies to eradicate these lethal seeds of cancer. The combination of conventional cancer therapies together with specific CSC targeted therapies brings the promise of eradicating a cancer with no possibility of recurrence. Yet, a multitude of surmountable barriers exist before CSC therapeutics can be fully deployed and clinical benefits derived. First and foremost, close cooperation between industry and academia is urgently needed to foster early stage drug discovery, biomarker characterization, unique animal model generation, target/pathway validation and clinical trial translation. As the field is less than a decade old, more basic CSC research is prerequisite to comprehensive understanding of the molecular basis of disease and the signaling pathways underlying each cancer and unique disease subtype. Another chief impediment to CSC therapeutic development is the exceptionally high cost to translate promising basic science discoveries in stem cell biology into meaningful clinical interventions. The lack of federal mechanisms large enough to support highly innovative research is at present a significant deterrent to translational science. New funding mechanisms must be developed to bring together academic researchers, disease foundations, donors, philanthropists, angel investors, government sponsors and industry stakeholders. Since existing RECIST criteria are not suitable endpoints for CSC trials [19], novel trial designs are vitally important at this juncture.

The most promising CSC therapeutic agents to date that target Notch, Hedgehog, Wnt, NFkB, Her2, Pten, Akt, Stat3, IL6R, CXCR1 and many other CSC self-renewal pathways afford ample opportunities for new partnerships and trials. Dual or multiple pathway inhibitors will be at a significant advantage since many cancers appear to be driven by multiple cellular regulators. Novel separation methods and characterization assays for better discrimination and quantification of CSCs are desperately needed. Imaging methods either ultrasonic, magnetic resonance or conventional microscopy-based are needed to better visualize CSCs before and after treatment to assess patient response. In addition, partnerships working toward CTC capture and characterization, nanoparticle delivery mechanisms and improvement of devices capable of sensitizing CSCs to chemotherapy and/or radiation clearance (e.g., low temperature hyperthermia induction) may be of great use in the fight against cancer in the near term. Similarly, business ventures that are able to develop clinical trial protocols capable of clearly elucidating CSC therapeutic targeting is urgently needed. Stem cell targeting approaches that can be co-administered with conventional debulking agents are most apt to quickly gain regulatory approval and yield the best results. Especially promising approaches would be those that take advantage of unique cell cycling aspects, apoptosis programs, epigenetics, miRNA regulation of EMT/MET states or other features of tumor initiating cells that would spare normal stem cells and the bodies’ innate regenerative capabilities.


  1. 1.
    Sternlicht MD et al (2006) Hormonal and local control of mammary branching morphogenesis. Differentiation 74(7):365–381PubMedCentralPubMedGoogle Scholar
  2. 2.
    Hinck L, Silberstein GB (2005) Key stages in mammary gland development: the mammary end bud as a motile organ. Breast Cancer Res 7(6):245–251PubMedCentralPubMedGoogle Scholar
  3. 3.
    Hennighausen L, Robinson GW (2005) Information networks in the mammary gland. Nat Rev Mol Cell Biol 6(9):715–725PubMedGoogle Scholar
  4. 4.
    Smalley M, Ashworth A (2003) Stem cells and breast cancer: a field in transit. Nat Rev Cancer 3(11):832–844PubMedGoogle Scholar
  5. 5.
    Kordon EC, Smith GH (1998) An entire functional mammary gland may comprise the progeny from a single cell. Development 125(10):1921–1930PubMedGoogle Scholar
  6. 6.
    Smith GH (1996) Experimental mammary epithelial morphogenesis in an in vivo model: evidence for distinct cellular progenitors of the ductal and lobular phenotype. Breast Cancer Res Treat 39(1):21–31PubMedGoogle Scholar
  7. 7.
    Asselin-Labat ML et al (2007) Gata-3 is an essential regulator of mammary-gland morphogenesis and luminal-cell differentiation. Nat Cell Biol 9(2):201–209PubMedGoogle Scholar
  8. 8.
    Stingl J et al (2006) Purification and unique properties of mammary epithelial stem cells. Nature 439(7079):993–997PubMedGoogle Scholar
  9. 9.
    Shackleton M et al (2006) Generation of a functional mammary gland from a single stem cell. Nature 439(7072):84–88PubMedGoogle Scholar
  10. 10.
    Sleeman KE et al (2006) CD24 staining of mouse mammary gland cells defines luminal epithelial, myoepithelial/basal and non-epithelial cells. Breast Cancer Res 8(1):R7PubMedCentralPubMedGoogle Scholar
  11. 11.
    Shipitsin M et al (2007) Molecular definition of breast tumor heterogeneity. Cancer Cell 11(3):259–273PubMedGoogle Scholar
  12. 12.
    Villadsen R et al (2007) Evidence for a stem cell hierarchy in the adult human breast. J Cell Biol 177(1):87–101PubMedCentralPubMedGoogle Scholar
  13. 13.
    Lim E et al (2009) Aberrant luminal progenitors as the candidate target population for basal tumor development in BRCA1 mutation carriers. Nat Med 15(8):907–913PubMedGoogle Scholar
  14. 14.
    Eirew P et al (2008) A method for quantifying normal human mammary epithelial stem cells with in vivo regenerative ability. Nat Med 14(12):1384–1389PubMedGoogle Scholar
  15. 15.
    Keller PJ et al (2011) Defining the cellular precursors to human breast cancer. Proc Natl Acad Sci U S A 109:2772–2777PubMedCentralPubMedGoogle Scholar
  16. 16.
    Visvader JE (2009) Keeping abreast of the mammary epithelial hierarchy and breast tumorigenesis. Genes Dev 23(22):2563–2577PubMedCentralPubMedGoogle Scholar
  17. 17.
    Bonnet D, Dick JE (1997) Human acute myeloid leukemia is organized as a hierarchy that originates from a primitive hematopoietic cell. Nat Med 3(7):730–737PubMedGoogle Scholar
  18. 18.
    Wicha MS, Liu S, Dontu G (2006) Cancer stem cells: an old idea—a paradigm shift. Cancer Res 66(4):1883–1890, discussion 1895-6PubMedGoogle Scholar
  19. 19.
    Liu S, Wicha MS (2010) Targeting breast cancer stem cells. J Clin Oncol 28(25):4006–4012PubMedGoogle Scholar
  20. 20.
    Charafe-Jauffret E et al (2008) Cancer stem cells in breast: current opinion and future challenges. Pathobiology 75(2):75–84PubMedCentralPubMedGoogle Scholar
  21. 21.
    Al-Hajj M et al (2003) Prospective identification of tumorigenic breast cancer cells. Proc Natl Acad Sci U S A 100(7):3983–3988PubMedCentralPubMedGoogle Scholar
  22. 22.
    Ginestier C et al (2007) ALDH1 is a marker of normal and malignant human mammary stem cells and a predictor of poor clinical outcome. Cell Stem Cell 1(5):555–567PubMedCentralPubMedGoogle Scholar
  23. 23.
    Liu S et al (2014) Breast cancer stem cells transition between epithelial and mesenchymal states reflective of their normal counterparts. Stem Cell Reports 2(1):78–91PubMedCentralPubMedGoogle Scholar
  24. 24.
    Singh SK et al (2004) Identification of human brain tumour initiating cells. Nature 432(7015):396–401PubMedGoogle Scholar
  25. 25.
    Collins AT et al (2005) Prospective identification of tumorigenic prostate cancer stem cells. Cancer Res 65(23):10946–10951PubMedGoogle Scholar
  26. 26.
    Patrawala L et al (2006) Highly purified CD44+ prostate cancer cells from xenograft human tumors are enriched in tumorigenic and metastatic progenitor cells. Oncogene 25(12):1696–1708PubMedGoogle Scholar
  27. 27.
    O’Brien CA et al (2007) A human colon cancer cell capable of initiating tumour growth in immunodeficient mice. Nature 445(7123):106–110PubMedGoogle Scholar
  28. 28.
    Ricci-Vitiani L et al (2007) Identification and expansion of human colon-cancer-initiating cells. Nature 445(7123):111–115PubMedGoogle Scholar
  29. 29.
    Li C et al (2007) Identification of pancreatic cancer stem cells. Cancer Res 67(3):1030–1037PubMedGoogle Scholar
  30. 30.
    Ma S et al (2007) Identification and characterization of tumorigenic liver cancer stem/progenitor cells. Gastroenterology 132(7):2542–2556PubMedGoogle Scholar
  31. 31.
    Ma S et al (2010) MiR-130b Promotes CD133(+) liver tumor-initiating cell growth and self-renewal via tumor protein 53-induced nuclear protein 1. Cell Stem Cell 7(6):694–707PubMedGoogle Scholar
  32. 32.
    Kim CF et al (2005) Identification of bronchioalveolar stem cells in normal lung and lung cancer. Cell 121(6):823–835PubMedGoogle Scholar
  33. 33.
    Prince ME et al (2007) Identification of a subpopulation of cells with cancer stem cell properties in head and neck squamous cell carcinoma. Proc Natl Acad Sci U S A 104(3):973–978PubMedCentralPubMedGoogle Scholar
  34. 34.
    Mukherjee S (2010) The emperor of all maladies: a biography of cancer. 1st Scribner hardcover ed, vol 14. Scribner, New York, NY, p 571, 8 p. of platesGoogle Scholar
  35. 35.
    Huntly BJ, Gilliland DG (2005) Cancer biology: summing up cancer stem cells. Nature 435(7046):1169–1170PubMedGoogle Scholar
  36. 36.
    Sell S (2004) Stem cell origin of cancer and differentiation therapy. Crit Rev Oncol Hematol 51(1):1–28PubMedGoogle Scholar
  37. 37.
    Xu Q et al (2009) Isolation of tumour stem-like cells from benign tumours. Br J Cancer 101(2):303–311PubMedCentralPubMedGoogle Scholar
  38. 38.
    Clay MR et al (2010) Single-marker identification of head and neck squamous cell carcinoma cancer stem cells with aldehyde dehydrogenase. Head Neck 32(9):1195–1201PubMedCentralPubMedGoogle Scholar
  39. 39.
    Silva IA et al (2011) Aldehyde dehydrogenase in combination with CD133 defines angiogenic ovarian cancer stem cells that portend poor patient survival. Cancer Res 71(11):3991–4001PubMedCentralPubMedGoogle Scholar
  40. 40.
    Krishnamurthy S et al (2010) Endothelial cell-initiated signaling promotes the survival and self-renewal of cancer stem cells. Cancer Res 70(23):9969–9978PubMedCentralPubMedGoogle Scholar
  41. 41.
    Fang D et al (2005) A tumorigenic subpopulation with stem cell properties in melanomas. Cancer Res 65(20):9328–9337PubMedGoogle Scholar
  42. 42.
    Quintana E et al (2008) Efficient tumour formation by single human melanoma cells. Nature 456(7222):593–598PubMedCentralPubMedGoogle Scholar
  43. 43.
    Roesch A et al (2010) A temporarily distinct subpopulation of slow-cycling melanoma cells is required for continuous tumor growth. Cell 141(4):583–594PubMedCentralPubMedGoogle Scholar
  44. 44.
    Boiko AD et al (2010) Human melanoma-initiating cells express neural crest nerve growth factor receptor CD271. Nature 466(7302):133–137PubMedCentralPubMedGoogle Scholar
  45. 45.
    Luo Y et al (2012) ALDH1A isozymes are markers of human melanoma stem cells and potential therapeutic targets. Stem Cells 30(10):2100–2113PubMedCentralPubMedGoogle Scholar
  46. 46.
    Kelly PN et al (2007) Tumor growth need not be driven by rare cancer stem cells. Science 317(5836):337PubMedGoogle Scholar
  47. 47.
    Charafe-Jauffret E et al (2009) Breast cancer cell lines contain functional cancer stem cells with metastatic capacity and a distinct molecular signature. Cancer Res 69(4):1302–1313PubMedCentralPubMedGoogle Scholar
  48. 48.
    Charafe-Jauffret E et al (2013) ALDH1-positive cancer stem cells predict engraftment of primary breast tumors and are governed by a common stem cell program. Cancer Res 73(24):7290–7300PubMedGoogle Scholar
  49. 49.
    Pece S et al (2010) Biological and molecular heterogeneity of breast cancers correlates with their cancer stem cell content. Cell 140(1):62–73PubMedGoogle Scholar
  50. 50.
    Bunting KD (2002) ABC transporters as phenotypic markers and functional regulators of stem cells. Stem Cells 20(1):11–20PubMedGoogle Scholar
  51. 51.
    Hadnagy A et al (2006) SP analysis may be used to identify cancer stem cell populations. Exp Cell Res 312(19):3701–3710PubMedGoogle Scholar
  52. 52.
    Hirschmann-Jax C et al (2004) A distinct “side population” of cells with high drug efflux capacity in human tumor cells. Proc Natl Acad Sci U S A 101(39):14228–14233PubMedCentralPubMedGoogle Scholar
  53. 53.
    Kondo T, Setoguchi T, Taga T (2004) Persistence of a small subpopulation of cancer stem-like cells in the C6 glioma cell line. Proc Natl Acad Sci U S A 101(3):781–786PubMedCentralPubMedGoogle Scholar
  54. 54.
    Clarke RB et al (2005) A putative human breast stem cell population is enriched for steroid receptor-positive cells. Dev Biol 277(2):443–456PubMedGoogle Scholar
  55. 55.
    Alvi AJ et al (2003) Functional and molecular characterisation of mammary side population cells. Breast Cancer Res 5(1):R1–R8PubMedCentralPubMedGoogle Scholar
  56. 56.
    Clayton H, Titley I, Vivanco M (2004) Growth and differentiation of progenitor/stem cells derived from the human mammary gland. Exp Cell Res 297(2):444–460PubMedGoogle Scholar
  57. 57.
    Welm BE et al (2002) Sca-1(pos) cells in the mouse mammary gland represent an enriched progenitor cell population. Dev Biol 245(1):42–56PubMedGoogle Scholar
  58. 58.
    Patrawala L et al (2005) Side population is enriched in tumorigenic, stem-like cancer cells, whereas ABCG2+ and ABCG2- cancer cells are similarly tumorigenic. Cancer Res 65(14):6207–6219PubMedGoogle Scholar
  59. 59.
    Britton KM et al (2012) Breast cancer, side population cells and ABCG2 expression. Cancer Lett 323(1):97–105PubMedGoogle Scholar
  60. 60.
    Nakanishi T et al (2010) Side-population cells in luminal-type breast cancer have tumour-initiating cell properties, and are regulated by HER2 expression and signalling. Br J Cancer 102(5):815–826PubMedCentralPubMedGoogle Scholar
  61. 61.
    Mani SA et al (2008) The epithelial-mesenchymal transition generates cells with properties of stem cells. Cell 133(4):704–715PubMedCentralPubMedGoogle Scholar
  62. 62.
    Marotta LL et al (2011) The JAK2/STAT3 signaling pathway is required for growth of CD44CD24 stem cell-like breast cancer cells in human tumors. J Clin Invest 121(7):2723–2735PubMedCentralPubMedGoogle Scholar
  63. 63.
    Honeth G et al (2008) The CD44+/CD24- phenotype is enriched in basal-like breast tumors. Breast Cancer Res 10(3):R53PubMedCentralPubMedGoogle Scholar
  64. 64.
    Meyer MJ et al (2010) CD44posCD49fhiCD133/2hi defines xenograft-initiating cells in estrogen receptor-negative breast cancer. Cancer Res 70(11):4624–4633PubMedCentralPubMedGoogle Scholar
  65. 65.
    Friedrichs K et al (1995) High expression level of alpha 6 integrin in human breast carcinoma is correlated with reduced survival. Cancer Res 55(4):901–906PubMedGoogle Scholar
  66. 66.
    Lipscomb EA et al (2005) The alpha6beta4 integrin maintains the survival of human breast carcinoma cells in vivo. Cancer Res 65(23):10970–10976PubMedGoogle Scholar
  67. 67.
    Chute JP et al (2006) Inhibition of aldehyde dehydrogenase and retinoid signaling induces the expansion of human hematopoietic stem cells. Proc Natl Acad Sci U S A 103(31):11707–11712PubMedCentralPubMedGoogle Scholar
  68. 68.
    Storms RW et al (1999) Isolation of primitive human hematopoietic progenitors on the basis of aldehyde dehydrogenase activity. Proc Natl Acad Sci U S A 96(16):9118–9123PubMedCentralPubMedGoogle Scholar
  69. 69.
    Ludeman SM (1999) The chemistry of the metabolites of cyclophosphamide. Curr Pharm Des 5(8):627–643PubMedGoogle Scholar
  70. 70.
    Cheung AM et al (2007) Aldehyde dehydrogenase activity in leukemic blasts defines a subgroup of acute myeloid leukemia with adverse prognosis and superior NOD/SCID engrafting potential. Leukemia 21(7):1423–1430PubMedGoogle Scholar
  71. 71.
    Sullivan JP et al (2010) Aldehyde dehydrogenase activity selects for lung adenocarcinoma stem cells dependent on notch signaling. Cancer Res 70(23):9937–9948PubMedCentralPubMedGoogle Scholar
  72. 72.
    Huang EH et al (2009) Aldehyde dehydrogenase 1 is a marker for normal and malignant human colonic stem cells (SC) and tracks SC overpopulation during colon tumorigenesis. Cancer Res 69(8):3382–3389PubMedCentralPubMedGoogle Scholar
  73. 73.
    Carpentino JE et al (2009) Aldehyde dehydrogenase-expressing colon stem cells contribute to tumorigenesis in the transition from colitis to cancer. Cancer Res 69(20):8208–8215PubMedCentralPubMedGoogle Scholar
  74. 74.
    van den Hoogen C et al (2010) High aldehyde dehydrogenase activity identifies tumor-initiating and metastasis-initiating cells in human prostate cancer. Cancer Res 70(12):5163–5173PubMedGoogle Scholar
  75. 75.
    Dontu G et al (2003) In vitro propagation and transcriptional profiling of human mammary stem/progenitor cells. Genes Dev 17(10):1253–1270PubMedCentralPubMedGoogle Scholar
  76. 76.
    D’Angelo RC, Wicha MS (2010) Stem cells in normal development and cancer. Prog Mol Biol Transl Sci 95:113–158PubMedGoogle Scholar
  77. 77.
    Kusumbe AP, Bapat SA (2009) Cancer stem cells and aneuploid populations within developing tumors are the major determinants of tumor dormancy. Cancer Res 69(24):9245–9253PubMedGoogle Scholar
  78. 78.
    Hendrikx PJ et al (1996) Homing of fluorescently labeled murine hematopoietic stem cells. Exp Hematol 24(2):129–140PubMedGoogle Scholar
  79. 79.
    Lanzkron SM, Collector MI, Sharkis SJ (1999) Homing of long-term and short-term engrafting cells in vivo. Ann N Y Acad Sci 872:48–54, discussion 54-6PubMedGoogle Scholar
  80. 80.
    Askenasy N, Farkas DL (2002) Optical imaging of PKH-labeled hematopoietic cells in recipient bone marrow in vivo. Stem Cells 20(6):501–513PubMedGoogle Scholar
  81. 81.
    Cicalese A et al (2009) The tumor suppressor p53 regulates polarity of self-renewing divisions in mammary stem cells. Cell 138(6):1083–1095PubMedGoogle Scholar
  82. 82.
    Lassailly F, Griessinger E, Bonnet D (2010) “Microenvironmental contaminations” induced by fluorescent lipophilic dyes used for noninvasive in vitro and in vivo cell tracking. Blood 115(26):5347–5354PubMedGoogle Scholar
  83. 83.
    Liu S, Clouthier SG, Wicha MS (2012) Role of microRNAs in the regulation of breast cancer stem cells. J Mammary Gland Biol Neoplasia 17(1):15–21PubMedCentralPubMedGoogle Scholar
  84. 84.
    Liu S et al (2012) MicroRNA93 regulates proliferation and differentiation of normal and malignant breast stem cells. PLoS Genet 8(6):e1002751PubMedCentralPubMedGoogle Scholar
  85. 85.
    Thiery JP (2003) Epithelial-mesenchymal transitions in development and pathologies. Curr Opin Cell Biol 15(6):740–746PubMedGoogle Scholar
  86. 86.
    Yang J et al (2004) Twist, a master regulator of morphogenesis, plays an essential role in tumor metastasis. Cell 117(7):927–939PubMedGoogle Scholar
  87. 87.
    Wu Y et al (2009) Stabilization of snail by NF-kappaB is required for inflammation-induced cell migration and invasion. Cancer Cell 15(5):416–428PubMedCentralPubMedGoogle Scholar
  88. 88.
    Yang MH et al (2008) Direct regulation of TWIST by HIF-1alpha promotes metastasis. Nat Cell Biol 10(3):295–305PubMedGoogle Scholar
  89. 89.
    Samavarchi-Tehrani P et al (2010) Functional genomics reveals a BMP-driven mesenchymal-to-epithelial transition in the initiation of somatic cell reprogramming. Cell Stem Cell 7(1):64–77PubMedGoogle Scholar
  90. 90.
    Brabletz T (2012) To differentiate or not: routes towards metastasis. Nat Rev Cancer 12(6):425–436PubMedGoogle Scholar
  91. 91.
    Tsai JH et al (2012) Spatiotemporal regulation of epithelial-mesenchymal transition is essential for squamous cell carcinoma metastasis. Cancer Cell 22(6):725–736PubMedCentralPubMedGoogle Scholar
  92. 92.
    Ocana OH et al (2012) Metastatic colonization requires the repression of the epithelial-mesenchymal transition inducer Prrx1. Cancer Cell 22(6):709–724PubMedGoogle Scholar
  93. 93.
    Malanchi I et al (2012) Interactions between cancer stem cells and their niche govern metastatic colonization. Nature 481(7379):85–89Google Scholar
  94. 94.
    Korpal M et al (2011) Direct targeting of Sec23a by miR-200s influences cancer cell secretome and promotes metastatic colonization. Nat Med 17(9):1101–1108PubMedCentralPubMedGoogle Scholar
  95. 95.
    Stankic M et al (2013) TGF-beta-Id1 signaling opposes Twist1 and promotes metastatic colonization via a mesenchymal-to-epithelial transition. Cell Rep 5(5):1228–1242PubMedCentralPubMedGoogle Scholar
  96. 96.
    Herschkowitz JI et al (2007) Identification of conserved gene expression features between murine mammary carcinoma models and human breast tumors. Genome Biol 8(5):R76PubMedCentralPubMedGoogle Scholar
  97. 97.
    Perou CM et al (2000) Molecular portraits of human breast tumours. Nature 406(6797):747–752PubMedGoogle Scholar
  98. 98.
    Sorlie T et al (2001) Gene expression patterns of breast carcinomas distinguish tumor subclasses with clinical implications. Proc Natl Acad Sci U S A 98(19):10869–10874PubMedCentralPubMedGoogle Scholar
  99. 99.
    Sotiriou C et al (2003) Breast cancer classification and prognosis based on gene expression profiles from a population-based study. Proc Natl Acad Sci U S A 100(18):10393–10398PubMedCentralPubMedGoogle Scholar
  100. 100.
    Chang JC et al (2005) Patterns of resistance and incomplete response to docetaxel by gene expression profiling in breast cancer patients. J Clin Oncol 23(6):1169–1177PubMedGoogle Scholar
  101. 101.
    Creighton CJ et al (2009) Residual breast cancers after conventional therapy display mesenchymal as well as tumor-initiating features. Proc Natl Acad Sci U S A 106(33):13820–13825PubMedCentralPubMedGoogle Scholar
  102. 102.
    Molyneux G et al (2010) BRCA1 basal-like breast cancers originate from luminal epithelial progenitors and not from basal stem cells. Cell Stem Cell 7(3):403–417PubMedGoogle Scholar
  103. 103.
    Proia TA et al (2011) Genetic predisposition directs breast cancer phenotype by dictating progenitor cell fate. Cell Stem Cell 8(2):149–163PubMedCentralPubMedGoogle Scholar
  104. 104.
    Keller PJ et al (2012) Defining the cellular precursors to human breast cancer. Proc Natl Acad Sci U S A 109(8):2772–2777PubMedCentralPubMedGoogle Scholar
  105. 105.
    Van Keymeulen A et al (2011) Distinct stem cells contribute to mammary gland development and maintenance. Nature 479(7372):189–193PubMedGoogle Scholar
  106. 106.
    Jeselsohn R et al (2010) Cyclin D1 kinase activity is required for the self-renewal of mammary stem and progenitor cells that are targets of MMTV-ErbB2 tumorigenesis. Cancer Cell 17(1):65–76PubMedCentralPubMedGoogle Scholar
  107. 107.
    Liu S et al (2008) BRCA1 regulates human mammary stem/progenitor cell fate. Proc Natl Acad Sci U S A 105(5):1680–1685PubMedCentralPubMedGoogle Scholar
  108. 108.
    Lagadec C et al (2010) Survival and self-renewing capacity of breast cancer initiating cells during fractionated radiation treatment. Breast Cancer Res 12(1):R13PubMedCentralPubMedGoogle Scholar
  109. 109.
    Phillips TM, McBride WH, Pajonk F (2006) The response of CD24(-/low)/CD44+ breast cancer-initiating cells to radiation. J Natl Cancer Inst 98(24):1777–1785PubMedGoogle Scholar
  110. 110.
    Karimi-Busheri F et al (2010) Senescence evasion by MCF-7 human breast tumor-initiating cells. Breast Cancer Res 12(3):R31PubMedCentralPubMedGoogle Scholar
  111. 111.
    Fillmore CM, Kuperwasser C (2008) Human breast cancer cell lines contain stem-like cells that self-renew, give rise to phenotypically diverse progeny and survive chemotherapy. Breast Cancer Res 10(2):R25PubMedCentralPubMedGoogle Scholar
  112. 112.
    Shafee N et al (2008) Cancer stem cells contribute to cisplatin resistance in Brca1/p53-mediated mouse mammary tumors. Cancer Res 68(9):3243–3250PubMedCentralPubMedGoogle Scholar
  113. 113.
    Woodward WA et al (2007) WNT/beta-catenin mediates radiation resistance of mouse mammary progenitor cells. Proc Natl Acad Sci U S A 104(2):618–623PubMedCentralPubMedGoogle Scholar
  114. 114.
    Diehn M et al (2009) Association of reactive oxygen species levels and radioresistance in cancer stem cells. Nature 458(7239):780–783PubMedCentralPubMedGoogle Scholar
  115. 115.
    Yu F et al (2007) let-7 regulates self renewal and tumorigenicity of breast cancer cells. Cell 131(6):1109–23PubMedGoogle Scholar
  116. 116.
    Zielske SP et al (2011) Ablation of breast cancer stem cells with radiation. Transl Oncol 4(4):227–233PubMedCentralPubMedGoogle Scholar
  117. 117.
    Li X et al (2008) Intrinsic resistance of tumorigenic breast cancer cells to chemotherapy. J Natl Cancer Inst 100(9):672–679PubMedGoogle Scholar
  118. 118.
    Korkaya H et al (2008) HER2 regulates the mammary stem/progenitor cell population driving tumorigenesis and invasion. Oncogene 27(47):6120–6130PubMedCentralPubMedGoogle Scholar
  119. 119.
    Tanei T et al (2009) Association of breast cancer stem cells identified by aldehyde dehydrogenase 1 expression with resistance to sequential Paclitaxel and epirubicin-based chemotherapy for breast cancers. Clin Cancer Res 15(12):4234–4241PubMedGoogle Scholar
  120. 120.
    Behbod F et al (2006) Transcriptional profiling of mammary gland side population cells. Stem Cells 24(4):1065–1074PubMedGoogle Scholar
  121. 121.
    Ithimakin S et al (2013) HER2 drives luminal breast cancer stem cells in the absence of HER2 amplification: implications for efficacy of adjuvant trastuzumab. Cancer Res 73(5):1635–1646PubMedCentralPubMedGoogle Scholar
  122. 122.
    Sladek NE (2003) Human aldehyde dehydrogenases: potential pathological, pharmacological, and toxicological impact. J Biochem Mol Toxicol 17(1):7–23PubMedGoogle Scholar
  123. 123.
    Su Y et al (2010) Aldehyde dehydrogenase 1 A1-positive cell population is enriched in tumor-initiating cells and associated with progression of bladder cancer. Cancer Epidemiol Biomarkers Prev 19(2):327–337PubMedCentralPubMedGoogle Scholar
  124. 124.
    Kim MP et al (2011) ALDH activity selectively defines an enhanced tumor-initiating cell population relative to CD133 expression in human pancreatic adenocarcinoma. PLoS One 6(6):e20636PubMedCentralPubMedGoogle Scholar
  125. 125.
    Landen CN Jr et al (2010) Targeting aldehyde dehydrogenase cancer stem cells in ovarian cancer. Mol Cancer Ther 9(12):3186–3199PubMedCentralPubMedGoogle Scholar
  126. 126.
    Magni M et al (1996) Induction of cyclophosphamide-resistance by aldehyde-dehydrogenase gene transfer. Blood 87(3):1097–1103PubMedGoogle Scholar
  127. 127.
    Moreb J et al (1996) Overexpression of the human aldehyde dehydrogenase class I results in increased resistance to 4-hydroperoxycyclophosphamide. Cancer Gene Ther 3(1):24–30PubMedGoogle Scholar
  128. 128.
    Moreb JS et al (2000) Expression of antisense RNA to aldehyde dehydrogenase class-1 sensitizes tumor cells to 4-hydroperoxycyclophosphamide in vitro. J Pharmacol Exp Ther 293(2):390–396PubMedGoogle Scholar
  129. 129.
    Sun QL et al (2011) Comparative proteomic analysis of paclitaxel sensitive A549 lung adenocarcinoma cell line and its resistant counterpart A549-Taxol. J Cancer Res Clin Oncol 137(3):521–532PubMedGoogle Scholar
  130. 130.
    Sladek NE et al (2002) Cellular levels of aldehyde dehydrogenases (ALDH1A1 and ALDH3A1) as predictors of therapeutic responses to cyclophosphamide-based chemotherapy of breast cancer: a retrospective study. Rational individualization of oxazaphosphorine-based cancer chemotherapeutic regimens. Cancer Chemother Pharmacol 49(4):309–321PubMedGoogle Scholar
  131. 131.
    Croker AK, Allan AL (2012) Inhibition of aldehyde dehydrogenase (ALDH) activity reduces chemotherapy and radiation resistance of stem-like ALDHhiCD44(+) human breast cancer cells. Breast Cancer Res Treat 133(1):75–87PubMedGoogle Scholar
  132. 132.
    Ward JF (1985) Biochemistry of DNA lesions. Radiat Res Suppl 8:S103–S111PubMedGoogle Scholar
  133. 133.
    Powell S, McMillan TJ (1990) DNA damage and repair following treatment with ionizing radiation. Radiother Oncol 19(2):95–108PubMedGoogle Scholar
  134. 134.
    Kryston TB et al (2011) Role of oxidative stress and DNA damage in human carcinogenesis. Mutat Res 711(1–2):193–201PubMedGoogle Scholar
  135. 135.
    Smith J et al (2000) Redox state is a central modulator of the balance between self-renewal and differentiation in a dividing glial precursor cell. Proc Natl Acad Sci U S A 97(18):10032–10037PubMedCentralPubMedGoogle Scholar
  136. 136.
    Ito K et al (2004) Regulation of oxidative stress by ATM is required for self-renewal of haematopoietic stem cells. Nature 431(7011):997–1002PubMedGoogle Scholar
  137. 137.
    Ito K et al (2006) Reactive oxygen species act through p38 MAPK to limit the lifespan of hematopoietic stem cells. Nat Med 12(4):446–451PubMedGoogle Scholar
  138. 138.
    Tothova Z et al (2007) FoxOs are critical mediators of hematopoietic stem cell resistance to physiologic oxidative stress. Cell 128(2):325–339PubMedGoogle Scholar
  139. 139.
    Miyamoto K et al (2007) Foxo3a is essential for maintenance of the hematopoietic stem cell pool. Cell Stem Cell 1(1):101–112PubMedGoogle Scholar
  140. 140.
    Guzman ML et al (2005) The sesquiterpene lactone parthenolide induces apoptosis of human acute myelogenous leukemia stem and progenitor cells. Blood 105(11):4163–4169PubMedCentralPubMedGoogle Scholar
  141. 141.
    Bao S et al (2006) Glioma stem cells promote radioresistance by preferential activation of the DNA damage response. Nature 444(7120):756–760PubMedGoogle Scholar
  142. 142.
    Yin H, Glass J (2011) The phenotypic radiation resistance of CD44+/CD24(-or low) breast cancer cells is mediated through the enhanced activation of ATM signaling. PLoS One 6(9):e24080PubMedCentralPubMedGoogle Scholar
  143. 143.
    Zhang M, Atkinson RL, Rosen JM (2010) Selective targeting of radiation-resistant tumor-initiating cells. Proc Natl Acad Sci U S A 107(8):3522–3527PubMedCentralPubMedGoogle Scholar
  144. 144.
    Ponta H, Sherman L, Herrlich PA (2003) CD44: from adhesion molecules to signalling regulators. Nat Rev Mol Cell Biol 4(1):33–45PubMedGoogle Scholar
  145. 145.
    Cheng C, Yaffe MB, Sharp PA (2006) A positive feedback loop couples Ras activation and CD44 alternative splicing. Genes Dev 20(13):1715–1720PubMedCentralPubMedGoogle Scholar
  146. 146.
    Orian-Rousseau V et al (2002) CD44 is required for two consecutive steps in HGF/c-Met signaling. Genes Dev 16(23):3074–3086PubMedCentralPubMedGoogle Scholar
  147. 147.
    Sherman LS et al (2000) CD44 enhances neuregulin signaling by Schwann cells. J Cell Biol 150(5):1071–1084PubMedCentralPubMedGoogle Scholar
  148. 148.
    Bourguignon LY et al (1997) Interaction between the adhesion receptor, CD44, and the oncogene product, p185HER2, promotes human ovarian tumor cell activation. J Biol Chem 272(44):27913–27918PubMedGoogle Scholar
  149. 149.
    Draffin JE et al (2004) CD44 potentiates the adherence of metastatic prostate and breast cancer cells to bone marrow endothelial cells. Cancer Res 64(16):5702–5711PubMedGoogle Scholar
  150. 150.
    Hiraga T, Ito S, Nakamura H (2013) Cancer stem-like cell marker CD44 promotes bone metastases by enhancing tumorigenicity, cell motility, and hyaluronan production. Cancer Res 73(13):4112–4122PubMedGoogle Scholar
  151. 151.
    Brown RL et al (2011) CD44 splice isoform switching in human and mouse epithelium is essential for epithelial-mesenchymal transition and breast cancer progression. J Clin Invest 121(3):1064–1074PubMedCentralPubMedGoogle Scholar
  152. 152.
    Kouros-Mehr H et al (2008) GATA-3 links tumor differentiation and dissemination in a luminal breast cancer model. Cancer Cell 13(2):141–152PubMedCentralPubMedGoogle Scholar
  153. 153.
    Schabath H et al (2006) CD24 affects CXCR4 function in pre-B lymphocytes and breast carcinoma cells. J Cell Sci 119(Pt 2):314–325PubMedGoogle Scholar
  154. 154.
    Cojoc M et al (2013) Emerging targets in cancer management: role of the CXCL12/CXCR4 axis. Onco Targets Ther 6:1347–1361PubMedCentralPubMedGoogle Scholar
  155. 155.
    Sheridan C et al (2006) CD44+/CD24− breast cancer cells exhibit enhanced invasive properties: an early step necessary for metastasis. Breast Cancer Res 8(5):R59PubMedCentralPubMedGoogle Scholar
  156. 156.
    Liu R et al (2007) The prognostic role of a gene signature from tumorigenic breast-cancer cells. N Engl J Med 356(3):217–226PubMedGoogle Scholar
  157. 157.
    Balic M et al (2006) Most early disseminated cancer cells detected in bone marrow of breast cancer patients have a putative breast cancer stem cell phenotype. Clin Cancer Res 12(19):5615–5621PubMedGoogle Scholar
  158. 158.
    Liu H et al (2010) Cancer stem cells from human breast tumors are involved in spontaneous metastases in orthotopic mouse models. Proc Natl Acad Sci U S A 107(42):18115–18120PubMedCentralPubMedGoogle Scholar
  159. 159.
    Theodoropoulos PA et al (2010) Circulating tumor cells with a putative stem cell phenotype in peripheral blood of patients with breast cancer. Cancer Lett 288(1):99–106PubMedGoogle Scholar
  160. 160.
    Charafe-Jauffret E et al (2010) Aldehyde dehydrogenase 1-positive cancer stem cells mediate metastasis and poor clinical outcome in inflammatory breast cancer. Clin Cancer Res 16(1):45–55PubMedCentralPubMedGoogle Scholar
  161. 161.
    Liu S et al (2011) Breast cancer stem cells are regulated by mesenchymal stem cells through cytokine networks. Cancer Res 71(2):614–624PubMedCentralPubMedGoogle Scholar
  162. 162.
    Lianidou ES, Markou A (2011) Circulating tumor cells in breast cancer: detection systems, molecular characterization, and future challenges. Clin Chem 57(9):1242–1255PubMedGoogle Scholar
  163. 163.
    Murray NP et al (2012) Redefining micrometastasis in prostate cancer: a comparison of circulating prostate cells, bone marrow disseminated tumor cells and micrometastasis—implications in determining local or systemic treatment for biochemical failure after radical prostatectomy. Int J Mol Med 30(4):896–904PubMedGoogle Scholar
  164. 164.
    Cristofanilli M et al (2005) Circulating tumor cells: a novel prognostic factor for newly diagnosed metastatic breast cancer. J Clin Oncol 23(7):1420–1430PubMedGoogle Scholar
  165. 165.
    Joosse SA, Pantel K (2013) Biologic challenges in the detection of circulating tumor cells. Cancer Res 73(1):8–11PubMedGoogle Scholar
  166. 166.
    Maheswaran S et al (2008) Detection of mutations in EGFR in circulating lung-cancer cells. N Engl J Med 359(4):366–377PubMedCentralPubMedGoogle Scholar
  167. 167.
    Nagrath S et al (2007) Isolation of rare circulating tumour cells in cancer patients by microchip technology. Nature 450(7173):1235–1239PubMedCentralPubMedGoogle Scholar
  168. 168.
    Baccelli I et al (2013) Identification of a population of blood circulating tumor cells from breast cancer patients that initiates metastasis in a xenograft assay. Nat Biotechnol 31(6):539–544PubMedGoogle Scholar
  169. 169.
    Zhang L et al (2013) The identification and characterization of breast cancer CTCs competent for brain metastasis. Sci Transl Med 5(180):180ra48PubMedGoogle Scholar
  170. 170.
    Yu M et al (2013) Circulating breast tumor cells exhibit dynamic changes in epithelial and mesenchymal composition. Science 339(6119):580–584PubMedCentralPubMedGoogle Scholar
  171. 171.
    Pestrin M et al (2012) Final results of a multicenter phase II clinical trial evaluating the activity of single-agent lapatinib in patients with HER2-negative metastatic breast cancer and HER2-positive circulating tumor cells. A proof-of-concept study. Breast Cancer Res Treat 134(1):283–289PubMedGoogle Scholar
  172. 172.
    Giordano A et al (2011) Artificial neural network analysis of circulating tumor cells in metastatic breast cancer patients. Breast Cancer Res Treat 129(2):451–458PubMedGoogle Scholar
  173. 173.
    Cohen SJ et al (2009) Prognostic significance of circulating tumor cells in patients with metastatic colorectal cancer. Ann Oncol 20(7):1223–1229PubMedGoogle Scholar
  174. 174.
    Olmos D et al (2009) Circulating tumour cell (CTC) counts as intermediate end points in castration-resistant prostate cancer (CRPC): a single-centre experience. Ann Oncol 20(1):27–33PubMedGoogle Scholar
  175. 175.
    Allard WJ et al (2004) Tumor cells circulate in the peripheral blood of all major carcinomas but not in healthy subjects or patients with nonmalignant diseases. Clin Cancer Res 10(20):6897–6904PubMedGoogle Scholar
  176. 176.
    Whitesides GM (2006) The origins and the future of microfluidics. Nature 442(7101):368–373PubMedGoogle Scholar
  177. 177.
    Khandurina J et al (2000) Integrated system for rapid PCR-based DNA analysis in microfluidic devices. Anal Chem 72(13):2995–3000PubMedGoogle Scholar
  178. 178.
    Erickson D et al (2004) Electrokinetically controlled DNA hybridization microfluidic chip enabling rapid target analysis. Anal Chem 76(24):7269–7277PubMedGoogle Scholar
  179. 179.
    Haeberle S, Zengerle R (2007) Microfluidic platforms for lab-on-a-chip applications. Lab Chip 7(9):1094–1110PubMedGoogle Scholar
  180. 180.
    Gleghorn JP et al (2010) Capture of circulating tumor cells from whole blood of prostate cancer patients using geometrically enhanced differential immunocapture (GEDI) and a prostate-specific antibody. Lab Chip 10(1):27–29PubMedCentralPubMedGoogle Scholar
  181. 181.
    Adams AA et al (2008) Highly efficient circulating tumor cell isolation from whole blood and label-free enumeration using polymer-based microfluidics with an integrated conductivity sensor. J Am Chem Soc 130(27):8633–8641PubMedCentralPubMedGoogle Scholar
  182. 182.
    Stott SL et al (2010) Isolation of circulating tumor cells using a microvortex-generating herringbone-chip. Proc Natl Acad Sci U S A 107(43):18392–18397PubMedCentralPubMedGoogle Scholar
  183. 183.
    Yoon HJ et al (2013) Sensitive capture of circulating tumour cells by functionalized graphene oxide nanosheets. Nat Nanotechnol 8(10):735–741PubMedCentralPubMedGoogle Scholar
  184. 184.
    Lin HK et al (2010) Portable filter-based microdevice for detection and characterization of circulating tumor cells. Clin Cancer Res 16(20):5011–5018PubMedCentralPubMedGoogle Scholar
  185. 185.
    Vona G et al (2004) Impact of cytomorphological detection of circulating tumor cells in patients with liver cancer. Hepatology 39(3):792–797PubMedGoogle Scholar
  186. 186.
    Zheng S et al (2007) Membrane microfilter device for selective capture, electrolysis and genomic analysis of human circulating tumor cells. J Chromatogr A 1162(2):154–161PubMedGoogle Scholar
  187. 187.
    Zheng S et al (2011) 3D microfilter device for viable circulating tumor cell (CTC) enrichment from blood. Biomed Microdevices 13(1):203–213PubMedGoogle Scholar
  188. 188.
    Kuo JS et al (2010) Deformability considerations in filtration of biological cells. Lab Chip 10(7):837–842PubMedGoogle Scholar
  189. 189.
    Zhou J, Papautsky I (2013) Fundamentals of inertial focusing in microchannels. Lab Chip 13(6):1121–1132PubMedGoogle Scholar
  190. 190.
    Bhagat AA, Kuntaegowdanahalli SS, Papautsky I (2008) Continuous particle separation in spiral microchannels using Dean flows and differential migration. Lab Chip 8(11):1906–1914PubMedGoogle Scholar
  191. 191.
    Hou HW et al (2013) Isolation and retrieval of circulating tumor cells using centrifugal forces. Sci Rep 3:1259PubMedCentralPubMedGoogle Scholar
  192. 192.
    Ozkumur E et al (2013) Inertial focusing for tumor antigen-dependent and -independent sorting of rare circulating tumor cells. Sci Transl Med 5(179):179ra47PubMedCentralPubMedGoogle Scholar
  193. 193.
    Zhang W et al (2012) Microfluidics separation reveals the stem-cell-like deformability of tumor-initiating cells. Proc Natl Acad Sci U S A 109(46):18707–18712PubMedCentralPubMedGoogle Scholar
  194. 194.
    Kuperwasser C et al (2004) Reconstruction of functionally normal and malignant human breast tissues in mice. Proc Natl Acad Sci U S A 101(14):4966–4971PubMedCentralPubMedGoogle Scholar
  195. 195.
    DeRose YS et al (2011) Tumor grafts derived from women with breast cancer authentically reflect tumor pathology, growth, metastasis and disease outcomes. Nat Med 17(11):1514–1520PubMedCentralPubMedGoogle Scholar
  196. 196.
    Zhang X et al (2013) A renewable tissue resource of phenotypically stable, biologically and ethnically diverse, patient-derived human breast cancer xenograft models. Cancer Res 73(15):4885–4897PubMedCentralPubMedGoogle Scholar
  197. 197.
    Marangoni E et al (2007) A new model of patient tumor-derived breast cancer xenografts for preclinical assays. Clin Cancer Res 13(13):3989–3998PubMedGoogle Scholar
  198. 198.
    Beckhove P et al (2003) Efficient engraftment of human primary breast cancer transplants in nonconditioned NOD/Scid mice. Int J Cancer 105(4):444–453PubMedGoogle Scholar
  199. 199.
    Bergamaschi A et al (2009) Molecular profiling and characterization of luminal-like and basal-like in vivo breast cancer xenograft models. Mol Oncol 3(5–6):469–482PubMedGoogle Scholar
  200. 200.
    du Manoir S et al (2013) Breast tumor PDXs are genetically plastic and correspond to a subset of aggressive cancers prone to relapse. Mol Oncol 8:431–43PubMedGoogle Scholar
  201. 201.
    Clarke R et al (2001) Cellular and molecular pharmacology of antiestrogen action and resistance. Pharmacol Rev 53(1):25–71PubMedGoogle Scholar
  202. 202.
    Cottu P et al (2012) Modeling of response to endocrine therapy in a panel of human luminal breast cancer xenografts. Breast Cancer Res Treat 133(2):595–606PubMedGoogle Scholar
  203. 203.
    Kabos P et al (2012) Patient-derived luminal breast cancer xenografts retain hormone receptor heterogeneity and help define unique estrogen-dependent gene signatures. Breast Cancer Res Treat 135(2):415–432PubMedGoogle Scholar
  204. 204.
    Ginestier C et al (2010) CXCR1 blockade selectively targets human breast cancer stem cells in vitro and in xenografts. J Clin Invest 120(2):485–497PubMedCentralPubMedGoogle Scholar
  205. 205.
    Turkson J, Jove R (2000) STAT proteins: novel molecular targets for cancer drug discovery. Oncogene 19(56):6613–6626PubMedGoogle Scholar
  206. 206.
    Redell MS, Tweardy DJ (2005) Targeting transcription factors for cancer therapy. Curr Pharm Des 11(22):2873–2887PubMedGoogle Scholar
  207. 207.
    Chen Z, Han ZC (2008) STAT3: a critical transcription activator in angiogenesis. Med Res Rev 28(2):185–200PubMedGoogle Scholar
  208. 208.
    Dave B et al (2012) Selective small molecule Stat3 inhibitor reduces breast cancer tumor-initiating cells and improves recurrence free survival in a human-xenograft model. PLoS One 7(8):e30207PubMedCentralPubMedGoogle Scholar
  209. 209.
    Noguera-Troise I et al (2006) Blockade of Dll4 inhibits tumour growth by promoting non-productive angiogenesis. Nature 444(7122):1032–1037PubMedGoogle Scholar
  210. 210.
    Ridgway J et al (2006) Inhibition of Dll4 signalling inhibits tumour growth by deregulating angiogenesis. Nature 444(7122):1083–1087PubMedGoogle Scholar
  211. 211.
    Scehnet JS et al (2007) Inhibition of Dll4-mediated signaling induces proliferation of immature vessels and results in poor tissue perfusion. Blood 109(11):4753–4760PubMedCentralPubMedGoogle Scholar
  212. 212.
    Hoey T et al (2009) DLL4 blockade inhibits tumor growth and reduces tumor-initiating cell frequency. Cell Stem Cell 5(2):168–177PubMedGoogle Scholar
  213. 213.
    Schott AF et al (2013) Preclinical and clinical studies of gamma secretase inhibitors with docetaxel on human breast tumors. Clin Cancer Res 19(6):1512–1524PubMedCentralPubMedGoogle Scholar
  214. 214.
    Todaro M et al (2013) Erythropoietin activates cell survival pathways in breast cancer stem-like cells to protect them from chemotherapy. Cancer Res 73(21):6393–6400PubMedGoogle Scholar
  215. 215.
    Ingham PW, McMahon AP (2001) Hedgehog signaling in animal development: paradigms and principles. Genes Dev 15(23):3059–3087PubMedGoogle Scholar
  216. 216.
    Liu S et al (2006) Hedgehog signaling and Bmi-1 regulate self-renewal of normal and malignant human mammary stem cells. Cancer Res 66(12):6063–6071PubMedCentralPubMedGoogle Scholar
  217. 217.
    Kubo M et al (2004) Hedgehog signaling pathway is a new therapeutic target for patients with breast cancer. Cancer Res 64(17):6071–6074PubMedGoogle Scholar
  218. 218.
    Tremblay MR et al (2008) Semisynthetic cyclopamine analogues as potent and orally bioavailable hedgehog pathway antagonists. J Med Chem 51(21):6646–6649PubMedGoogle Scholar
  219. 219.
    Robarge KD et al (2009) GDC-0449-a potent inhibitor of the hedgehog pathway. Bioorg Med Chem Lett 19(19):5576–5581PubMedGoogle Scholar
  220. 220.
    Olive KP et al (2009) Inhibition of Hedgehog signaling enhances delivery of chemotherapy in a mouse model of pancreatic cancer. Science 324(5933):1457–1461PubMedCentralPubMedGoogle Scholar
  221. 221.
    Chiba S (2006) Notch signaling in stem cell systems. Stem Cells 24(11):2437–2447PubMedGoogle Scholar
  222. 222.
    Roy M, Pear WS, Aster JC (2007) The multifaceted role of Notch in cancer. Curr Opin Genet Dev 17(1):52–59PubMedGoogle Scholar
  223. 223.
    Radtke F, Raj K (2003) The role of Notch in tumorigenesis: oncogene or tumour suppressor? Nat Rev Cancer 3(10):756–767PubMedGoogle Scholar
  224. 224.
    Jhappan C et al (1992) Expression of an activated Notch-related int-3 transgene interferes with cell differentiation and induces neoplastic transformation in mammary and salivary glands. Genes Dev 6(3):345–355PubMedGoogle Scholar
  225. 225.
    Bouras T et al (2008) Notch signaling regulates mammary stem cell function and luminal cell-fate commitment. Cell Stem Cell 3(4):429–441PubMedGoogle Scholar
  226. 226.
    Reedijk M et al (2005) High-level coexpression of JAG1 and NOTCH1 is observed in human breast cancer and is associated with poor overall survival. Cancer Res 65(18):8530–8537PubMedGoogle Scholar
  227. 227.
    Harrison H et al (2010) Regulation of breast cancer stem cell activity by signaling through the Notch4 receptor. Cancer Res 70(2):709–718PubMedCentralPubMedGoogle Scholar
  228. 228.
    Zang S et al (2010) RNAi-mediated knockdown of Notch-1 leads to cell growth inhibition and enhanced chemosensitivity in human breast cancer. Oncol Rep 23(4):893–899PubMedGoogle Scholar
  229. 229.
    Olsauskas-Kuprys R, Zlobin A, Osipo C (2013) Gamma secretase inhibitors of Notch signaling. Onco Targets Ther 6:943–955PubMedCentralPubMedGoogle Scholar
  230. 230.
    Lindsay J et al (2008) ErbB2 induces Notch1 activity and function in breast cancer cells. Clin Transl Sci 1(2):107–115PubMedCentralPubMedGoogle Scholar
  231. 231.
    Farnie G et al (2013) Combined inhibition of ErbB1/2 and Notch receptors effectively targets breast ductal carcinoma in situ (DCIS) stem/progenitor cell activity regardless of ErbB2 status. PLoS One 8(2):e56840PubMedCentralPubMedGoogle Scholar
  232. 232.
    Osipo C et al (2008) ErbB-2 inhibition activates Notch-1 and sensitizes breast cancer cells to a gamma-secretase inhibitor. Oncogene 27(37):5019–5032PubMedGoogle Scholar
  233. 233.
    Won HY et al (2012) Loss of Mel-18 enhances breast cancer stem cell activity and tumorigenicity through activating Notch signaling mediated by the Wnt/TCF pathway. FASEB J 26(12):5002–5013PubMedGoogle Scholar
  234. 234.
    Rexer BN, Arteaga CL (2013) Optimal targeting of HER2-PI3K signaling in breast cancer: mechanistic insights and clinical implications. Cancer Res 73(13):3817–3820PubMedCentralPubMedGoogle Scholar
  235. 235.
    Nagata Y et al (2004) PTEN activation contributes to tumor inhibition by trastuzumab, and loss of PTEN predicts trastuzumab resistance in patients. Cancer Cell 6(2):117–127PubMedGoogle Scholar
  236. 236.
    Korkaya H et al (2009) Regulation of mammary stem/progenitor cells by PTEN/Akt/beta-catenin signaling. PLoS Biol 7(6):e1000121PubMedCentralPubMedGoogle Scholar
  237. 237.
    Junttila TT et al (2009) Ligand-independent HER2/HER3/PI3K complex is disrupted by trastuzumab and is effectively inhibited by the PI3K inhibitor GDC-0941. Cancer Cell 15(5):429–440PubMedGoogle Scholar
  238. 238.
    Serra V et al (2008) NVP-BEZ235, a dual PI3K/mTOR inhibitor, prevents PI3K signaling and inhibits the growth of cancer cells with activating PI3K mutations. Cancer Res 68(19):8022–8030PubMedGoogle Scholar
  239. 239.
    Andre F et al (2010) Phase I study of everolimus plus weekly paclitaxel and trastuzumab in patients with metastatic breast cancer pretreated with trastuzumab. J Clin Oncol 28(34):5110–5115PubMedGoogle Scholar
  240. 240.
    Chakrabarty A et al (2013) Trastuzumab-resistant cells rely on a HER2-PI3K-FoxO-survivin axis and are sensitive to PI3K inhibitors. Cancer Res 73(3):1190–1200PubMedCentralPubMedGoogle Scholar
  241. 241.
    Klaus A, Birchmeier W (2008) Wnt signalling and its impact on development and cancer. Nat Rev Cancer 8(5):387–398PubMedGoogle Scholar
  242. 242.
    Logan CY, Nusse R (2004) The Wnt signaling pathway in development and disease. Annu Rev Cell Dev Biol 20:781–810PubMedGoogle Scholar
  243. 243.
    Badders NM et al (2009) The Wnt receptor, Lrp5, is expressed by mouse mammary stem cells and is required to maintain the basal lineage. PLoS One 4(8):e6594PubMedCentralPubMedGoogle Scholar
  244. 244.
    Bafico A et al (2004) An autocrine mechanism for constitutive Wnt pathway activation in human cancer cells. Cancer Cell 6(5):497–506PubMedGoogle Scholar
  245. 245.
    Klopocki E et al (2004) Loss of SFRP1 is associated with breast cancer progression and poor prognosis in early stage tumors. Int J Oncol 25(3):641–649PubMedGoogle Scholar
  246. 246.
    Nagahata T et al (2003) Amplification, up-regulation and over-expression of DVL-1, the human counterpart of the Drosophila disheveled gene, in primary breast cancers. Cancer Sci 94(6):515–518PubMedGoogle Scholar
  247. 247.
    Nakopoulou L et al (2006) Study of phospho-beta-catenin subcellular distribution in invasive breast carcinomas in relation to their phenotype and the clinical outcome. Mod Pathol 19(4):556–563PubMedGoogle Scholar
  248. 248.
    Piva et al (2014) Sox2 promotes tamoxifen resistance in breast cancer cells. EMBO Mol Med 6:66–79Google Scholar
  249. 249.
    Li Y et al (2003) Evidence that transgenes encoding components of the Wnt signaling pathway preferentially induce mammary cancers from progenitor cells. Proc Natl Acad Sci U S A 100(26):15853–15858PubMedCentralPubMedGoogle Scholar
  250. 250.
    Liu BY et al (2004) The transforming activity of Wnt effectors correlates with their ability to induce the accumulation of mammary progenitor cells. Proc Natl Acad Sci U S A 101(12):4158–4163PubMedCentralPubMedGoogle Scholar
  251. 251.
    Conley SJ et al (2012) Antiangiogenic agents increase breast cancer stem cells via the generation of tumor hypoxia. Proc Natl Acad Sci U S A 109(8):2784–2789PubMedCentralPubMedGoogle Scholar
  252. 252.
    Yang ZQ et al (2009) Methylation-associated silencing of SFRP1 with an 8p11-12 amplification inhibits canonical and non-canonical WNT pathways in breast cancers. Int J Cancer 125(7):1613–1621PubMedCentralPubMedGoogle Scholar
  253. 253.
    He B et al (2004) A monoclonal antibody against Wnt-1 induces apoptosis in human cancer cells. Neoplasia 6(1):7–14PubMedCentralPubMedGoogle Scholar
  254. 254.
    DeAlmeida VI et al (2007) The soluble wnt receptor Frizzled8CRD-hFc inhibits the growth of teratocarcinomas in vivo. Cancer Res 67(11):5371–5379PubMedGoogle Scholar
  255. 255.
    Albini A, Sporn MB (2007) The tumour microenvironment as a target for chemoprevention. Nat Rev Cancer 7(2):139–147PubMedGoogle Scholar
  256. 256.
    Karnoub AE et al (2007) Mesenchymal stem cells within tumour stroma promote breast cancer metastasis. Nature 449(7162):557–563PubMedGoogle Scholar
  257. 257.
    Asselin-Labat ML et al (2010) Control of mammary stem cell function by steroid hormone signalling. Nature 465(7299):798–802PubMedGoogle Scholar
  258. 258.
    Fillmore CM et al (2010) Estrogen expands breast cancer stem-like cells through paracrine FGF/Tbx3 signaling. Proc Natl Acad Sci U S A 107(50):21737–21742PubMedCentralPubMedGoogle Scholar
  259. 259.
    Pierce BL et al (2009) Elevated biomarkers of inflammation are associated with reduced survival among breast cancer patients. J Clin Oncol 27(21):3437–3444PubMedCentralPubMedGoogle Scholar
  260. 260.
    Coussens LM, Werb Z (2002) Inflammation and cancer. Nature 420(6917):860–867PubMedCentralPubMedGoogle Scholar
  261. 261.
    Sansone P et al (2007) IL-6 triggers malignant features in mammospheres from human ductal breast carcinoma and normal mammary gland. J Clin Invest 117(12):3988–4002PubMedCentralPubMedGoogle Scholar
  262. 262.
    Conze D et al (2001) Autocrine production of interleukin 6 causes multidrug resistance in breast cancer cells. Cancer Res 61(24):8851–8858PubMedGoogle Scholar
  263. 263.
    Iliopoulos D, Hirsch HA, Struhl K (2009) An epigenetic switch involving NF-kappaB, Lin28, Let-7 MicroRNA, and IL6 links inflammation to cell transformation. Cell 139(4):693–706PubMedCentralPubMedGoogle Scholar
  264. 264.
    Korkaya H et al (2012) Activation of an IL6 inflammatory loop mediates trastuzumab resistance in HER2+ breast cancer by expanding the cancer stem cell population. Mol Cell 47(4):570–584PubMedCentralPubMedGoogle Scholar
  265. 265.
    Germain D, Frank DA (2007) Targeting the cytoplasmic and nuclear functions of signal transducers and activators of transcription 3 for cancer therapy. Clin Cancer Res 13(19):5665–5669PubMedGoogle Scholar
  266. 266.
    Lin L et al (2013) Evaluation of STAT3 signaling in ALDH+ and ALDH+/CD44+/CD24− subpopulations of breast cancer cells. PLoS One 8(12):e82821PubMedCentralPubMedGoogle Scholar
  267. 267.
    Yu F et al (2010) Mir-30 reduction maintains self-renewal and inhibits apoptosis in breast tumor-initiating cells. Oncogene 29(29):4194–4204PubMedGoogle Scholar
  268. 268.
    Shimono Y et al (2009) Downregulation of miRNA-200c links breast cancer stem cells with normal stem cells. Cell 138(3):592–603PubMedCentralPubMedGoogle Scholar
  269. 269.
    World Health Organization (2012) Cancer Fact Sheet N297.
  270. 270.
    Phend C (2011) Breast cervical cancer kill 625,000 women each year. (3/25/12).
  271. 271.
    National Cancer Institute. Surveillance epidemiology and end results: SEER stat fact sheets. Accessed 31 Mar 12.
  272. 272.
    American Cancer Society. Cancer facts and figures 2012. Accessed 31 Mar 12.
  273. 273.
    Perez EA et al (2010) HER2 and chromosome 17 effect on patient outcome in the N9831 adjuvant trastuzumab trial. J Clin Oncol 28(28):4307–4315PubMedCentralPubMedGoogle Scholar
  274. 274.
    Recht A et al (1996) The sequencing of chemotherapy and radiation therapy after conservative surgery for early-stage breast cancer. N Engl J Med 334(21):1356–1361PubMedGoogle Scholar
  275. 275.
    Slamon DJ et al (2001) Use of chemotherapy plus a monoclonal antibody against HER2 for metastatic breast cancer that overexpresses HER2. N Engl J Med 344(11):783–792PubMedGoogle Scholar
  276. 276.
    Hedge SR, Sun W, Lynch JP (2008) Systemic and targeted therapy for advanced colon cancer. Expert Rev Gastroenterol Hepatol 2(1):135–149Google Scholar
  277. 277.
    Silvestri GA, Rivera MP (2005) Targeted therapy for the treatment of advanced non-small cell lung cancer: a review of the epidermal growth factor receptor antagonists. Chest 128(6):3975–3984PubMedGoogle Scholar
  278. 278.
    Sherbenou DW, Druker BJ (2007) Applying the discovery of the Philadelphia chromosome. J Clin Invest 117(8):2067–2074PubMedCentralPubMedGoogle Scholar
  279. 279.
    Sakariassen PO, Immervoll H, Chekenya M (2007) Cancer stem cells as mediators of treatment resistance in brain tumors: status and controversies. Neoplasia 9(11):882–892PubMedCentralPubMedGoogle Scholar
  280. 280.
    Brekelmans CT et al (2007) Tumour characteristics, survival and prognostic factors of hereditary breast cancer from BRCA2-, BRCA1- and non-BRCA1/2 families as compared to sporadic breast cancer cases. Eur J Cancer 43(5):867–876PubMedGoogle Scholar
  281. 281.
    Korkaya H, Wicha MS (2007) Selective targeting of cancer stem cells: a new concept in cancer therapeutics. BioDrugs 21(5):299–310PubMedGoogle Scholar
  282. 282.
    Houston S (2011) Stemming the tide.
  283. 283.
    Markey K (2009) Firms seek to prove cancer stem cell hypothesis. Bio Market Trends. 29(18).

Copyright information

© Springer Science+Business Media LLC New York 2015

Authors and Affiliations

  • Ming Luo
    • 1
  • Shawn G. Clouthier
    • 1
  • Yadwinder Deol
    • 1
  • Suling Liu
    • 2
  • Sunitha Nagrath
    • 3
  • Ebrahim Azizi
    • 1
  • Max S. Wicha
    • 1
    Email author
  1. 1.Department of Internal MedicineUniversity of MichiganAnn ArborUSA
  2. 2.School of Life SciencesUniversity of Science and Technology of ChinaHefeiChina
  3. 3.Department of Chemical EngineeringUniversity of MichiganAnn ArborUSA

Personalised recommendations