The small subpopulation of breast cancer cells that possess the capability for self-renewal and formation of secondary tumours that recapitulate the heterogeneity of the primary tumour are referred to as tumour-initiating cells or BCSCs (breast cancer stem cells). The hypoxic tumour microenvironment and chemotherapy actively induce the BCSC phenotype. HIFs (hypoxia-inducible factors) are required and molecular mechanisms by which they promote the BCSC phenotype have recently been delineated. HIF inhibitors block chemotherapy-induced enrichment of BCSCs, suggesting that their use may improve the response to chemotherapy and increase the survival of breast cancer patients.

INTRODUCTION

Forty thousand women with breast cancer die every year in the U.S.A. because cancer cells metastasize to secondary sites and fail to respond to chemotherapy. The vast majority of cancer cells within a primary breast tumour are capable of only a limited number of cell divisions and therefore do not have the potential to form a clinically relevant secondary tumour. The small subpopulation of cancer cells that possess the capability for self-renewal and to form secondary tumours that recapitulate the heterogeneity of the primary tumour are referred to as tumour-initiating cells or BCSCs (breast cancer stem cells) [1]. BCSCs undergo asymmetric division to yield one BCSC and one bulk cancer cell (Figure 1A). These BCSCs must be effectively targeted in order to eradicate cancer, but, relative to bulk cancer cells, BCSCs have increased resistance to chemotherapy. Worse yet, the tumour microenvironment and chemotherapy actively induce bulk cancer cells to assume the BCSC phenotype. Thus 40000 women with breast cancer die every year in the U.S.A. because of the failure of existing therapies to eradicate BCSCs.

Characterization of BCSCs

Figure 1
Characterization of BCSCs

(A) BCSCs are capable of asymmetric cell division, which results in the formation of a daughter BCSC (self-renewal) and a bulk breast cancer cell (differentiation). (B and C) Several methods are available for determining the abundance of BCSCs including the Aldefluor assay, which measures ALDH activity (B) and the mammosphere assay, which is based on the ability of BCSCs to generate clusters of cells in suspension culture (C).

Figure 1
Characterization of BCSCs

(A) BCSCs are capable of asymmetric cell division, which results in the formation of a daughter BCSC (self-renewal) and a bulk breast cancer cell (differentiation). (B and C) Several methods are available for determining the abundance of BCSCs including the Aldefluor assay, which measures ALDH activity (B) and the mammosphere assay, which is based on the ability of BCSCs to generate clusters of cells in suspension culture (C).

Breast cancers can be classified by clinical, molecular or pathological criteria. Clinically, breast cancers are divided into three groups, each of which receives different therapies. Cancers that express ER (oestrogen receptor) or PR (progesterone receptor), or both, are treated with an ER antagonist, such as tamoxifen, or an aromatase inhibitor, such as letrozole. Cancers with amplification of the gene encoding HER2 (human epidermal growth factor receptor 2) are treated with an anti-HER2 antibody, such as trastuzumab, or a receptor tyrosine kinase inhibitor, such as lapatinib. Cancers that lack ER/PR expression or HER2 amplification, which are known as TNBCs (triple-negative breast cancers), are not responsive to the targeted therapies described above and are treated with cytotoxic chemotherapy with a durable response rate of less than 20%, leading to significantly increased mortality compared with the other two patient groups. A molecular system of classification, based on the mRNA expression of a 50-gene signature known as the PAM50, divides breast cancers into luminal-type tumours, which overlap with the ER/PR-expressing tumours, HER2-type tumours, and basal-type tumours, which overlap with the TNBCs [2]. The basal-type breast cancers are characterized by the highest percentage of BCSCs. Finally, histopathology divides tumours into low-grade well-differentiated (good prognosis) and high-grade poorly differentiated (poor prognosis) subtypes, with poorly differentiated breast cancers containing a higher percentage of BCSCs [3].

ASSAYS FOR COUNTING BCSCs

There are several assays that are used to quantify BCSCs. The gold standard for BCSCs is the ability of a limiting number of cells to form a tumour after transplantation into immunodeficient mice. Thus stringent selection of BCSCs by one or more of the methods described below may result in tumour formation after injection of a 1000-fold fewer cells compared with BCSC-depleted populations [4].

In luminal-type breast cancers, the subpopulation of cells that express high levels of the cell-surface marker CD44 and low levels of CD24 are enriched for BCSCs [5], whereas the percentage of CD44+CD24 cancer cells in basal-type tumours [6], and, especially, in breast cancer cell lines derived from them [7], is greatly increased and is therefore a less useful marker of BCSCs since it is expressed by many bulk cancer cells as well as BCSCs.

A second marker of BCSCs is ALDH (aldehyde dehydrogenase) activity, which can be assayed using the commercially available Aldefluor assay (Figure 1B). Breast cancer cells are incubated with an ALDH substrate (BODIPY–aminoacetaldehyde) that is converted into a fluorescent product (BODIPY–aminoacetate), which can be detected by flow cytometry [4]. In one study, 6.1% of breast cancer cells were CD44+CD24, 4.3% were ALDH+ and 1.2% were ALDH+CD44+CD24; tumour formation was not observed after injection of as many as 50000 cells that were CD44+CD24 but not ALDH+, whereas tumour formation was observed after injection of 1500 cells that were ALDH+ but not CD44+CD24 or as few as 20 cells that were both ALDH+ and CD44+CD24 [4]. Overall survival was significantly decreased in patients with breast cancers that were scored ALDH+ by immunohistochemistry [4].

A third method of enriching for BCSCs is the mammosphere formation assay, which is based on the ability of BCSCs, but not bulk cancer cells, to propagate as multicellular spheroids in suspension culture [8]. The primary mammospheres can be harvested, dissociated, replated and assayed for secondary mammosphere formation as a more rigorous test of self-renewal capacity (Figure 1C). Mammospheres are also highly enriched in tumour-initiating cells. Each of these three methods enriches for BCSCs, but none of them has complete selectivity and sensitivity. As a result, demonstrating similar results using two different assays is most desirable.

INTRATUMORAL HYPOXIA DRIVES BREAST CANCER INVASION AND METASTASIS

Studies employing Eppendorf microelectrodes to directly measure tissue oxygenation revealed a median pO2 in locally advanced breast cancers of 10 mmHg, compared with 65 mmHg in normal breast tissue [9]. The reduced O2 availability that is commonly observed in advanced breast cancers is due to diffusion-limited hypoxia, which occurs when cells are too far away from a blood vessel to receive adequate oxygenation, and perfusion-limited hypoxia, which occurs due to failure of blood to flow through structurally and functionally abnormal tumour blood vessels [9]. Similar degrees of intratumoral hypoxia were reported in several other types of cancer, as was the finding that pO2<10 mmHg was associated with a significantly increased risk of patient mortality [10].

One of the major mechanisms by which intratumoral hypoxia influences the outcome of breast cancer pathogenesis is through the activation of HIFs (hypoxia-inducible factors), which are heterodimeric transcription factors that consist of an O2-regulated HIF-α subunit (HIF-1α, HIF-2α or HIF-3α) and a constitutively expressed HIF-1β subunit, as first demonstrated for HIF-1 [11,12]. Under aerobic conditions, the HIF-α subunits are degraded in a four-step process: first, they are subjected to O2-dependent prolyl hydroxylation; secondly, the hydroxylated proteins are bound by the VHL (von Hippel–Lindau) tumour-suppressor protein; thirdly, VHL recruits an E3 ubiquitin–protein ligase complex; and fourthly, the polyubiquitinated HIF-α proteins are subjected to proteasomal degradation [13]. In addition, HIF-1α and HIF-2α are subject to O2-dependent asparaginyl hydroxylation, which blocks their interaction with the co-activator proteins p300 and CBP. Under hypoxic conditions, prolyl and asparaginyl hydroxylation are inhibited and the HIF-α proteins rapidly accumulate, dimerize with HIF-1β, recruit co-activators and activate the transcription of specific target genes by binding to DNA at cis-acting hypoxia-response elements, which all contain the core sequence 5′-(A/G)CGTG-3′ [13]. Over 1500 HIF target genes have been identified, although, in any given cell, only a subset of target genes (several hundred) will be transactivated by HIFs in response to hypoxia. The expression of an approximately equal number of genes is decreased in response to hypoxia. Although repression of these genes is HIF-dependent, in most cases it is not mediated by direct HIF binding; instead, HIFs indirectly inhibit gene expression through regulation of genes that encode transcriptional repressors, chromatin-modifying enzymes, and miRNAs [14].

Given the severe hypoxia that is common in human breast cancers, it is not surprising that HIF-1α and HIF-2α protein levels were found by immunohistochemistry to be greatly increased in breast cancer biopsies compared with surrounding normal breast tissue [15,16]. Immunohistochemical analysis of HIF-1α protein levels in primary breast cancer biopsies has demonstrated a significant association with mortality in ten clinical studies that altogether include several thousand patients [1726]. A multitude of HIF-3α isoforms are generated by alternative splicing and, depending on the particular isoform, may function as an activator or repressor of HIF-dependent transcription [27,28]. The role of these isoforms in breast cancer progression has not been determined.

Pre-clinical studies in autochthonous (conditional knockout) and orthotopic mouse models have demonstrated that both HIF-1α and HIF-2α are required for breast cancer primary tumour growth and metastasis to lymph nodes and lungs [2931]. Conditional knockouts have also demonstrated that HIF activity plays important roles in tumour stromal cells [3234]. Specific HIF target genes have been identified that contribute to discrete steps in the metastatic process (Table 1). Other HIF target genes are required for metastasis, but function in a more indirect manner. For example, RAB22A encodes a protein that is required for hypoxia-induced microvesicle formation, which allows hypoxic breast cancer cells to export the invasive/metastatic phenotype to non-hypoxic cancer cells [35]. KDM4C is a HIF target gene that encodes a histone demethylase, which is itself recruited by HIF-1α to hypoxia-response elements where it increases HIF-1-dependent transactivation, and this feedforward mechanism is required for lung metastasis [36].

Table 1
HIF target genes that contribute to discrete steps in the metastatic process

ECM, extracellular matrix; MSC, mesenchymal stem cell.

Target gene(s)Metastatic processReference(s)
PDGFB Lymphangiogenesis, lymphatic invasion [30
CCR5, CSF1, CXCL16, CXCR3, PGF MSC and macrophage recruitment [73,74
RHOA, ROCK1 Cell motility [75
P4HA1, P4HA2, PLOD2 ECM remodelling, invasion, intravasation [76,77
L1CAM Margination [31
ANGPTL4 Extravasation [31
LOX, LOXL2, LOXL4 ECM remodelling, metastatic niche formation [7880
WWTR1, SIAH1 Cancer stem cell specification [51
Target gene(s)Metastatic processReference(s)
PDGFB Lymphangiogenesis, lymphatic invasion [30
CCR5, CSF1, CXCL16, CXCR3, PGF MSC and macrophage recruitment [73,74
RHOA, ROCK1 Cell motility [75
P4HA1, P4HA2, PLOD2 ECM remodelling, invasion, intravasation [76,77
L1CAM Margination [31
ANGPTL4 Extravasation [31
LOX, LOXL2, LOXL4 ECM remodelling, metastatic niche formation [7880
WWTR1, SIAH1 Cancer stem cell specification [51

INDUCTION OF THE BCSC PHENOTYPE BY HYPOXIA

Whereas many breast cancer cells may successfully invade the vasculature, only BCSCs are capable of forming a clinically relevant metastasis at a distant site [37]. In addition to triggering HIF-dependent transactivation of multiple genes required for invasion and metastasis, HIFs also regulate the BCSC phenotype. Exposure of multiple cancer cell lines, including MCF-7 breast cancer cells, to hypoxia revealed increased expression of the KLF4, NANOG, OCT4 and SOX2 genes, which encode pluripotency factors that are required for the maintenance of embryonic stem cells and induced pluripotent stem cells [38], although the molecular mechanisms underlying the increased expression were not determined. The first direct evidence for induction of the BCSC phenotype by intratumoral hypoxia was obtained by analysing tumour-bearing mice treated with anti-angiogenic agents [39]. HIF-1 was first shown to regulate angiogenesis through the identification of the VEGF gene, which encodes vascular endothelial growth factor, as a direct HIF-1 target gene [40]. VEGF plays a critical role in tumour vascularization, and bevacizumab, which is a monoclonal antibody that binds to VEGF, as well as small-molecule inhibitors of VEGF receptor tyrosine kinase activity, such as sunitinib, were shown to dramatically inhibit tumour growth and vascularization in mouse models [41].

Anti-angiogenic agents have had only modest success as cancer therapeutics, leading the U.S. FDA (Food and Drug Administration) to revoke its approval for use of bevacizumab in breast cancer patients [42]. Of even greater concern, pre-clinical studies revealed that the residual tumours in mice treated with anti-angiogenic agents had increased invasive and metastatic properties, which were attributed to the increased intratumoral hypoxia that resulted from inhibition of angiogenesis [43,44]. Analysis of mice bearing human breast tumour xenografts treated with sunitinib revealed an enrichment of BCSCs within hypoxic regions of the tumours [39]. Remarkably, exposure of human breast cancer cells to hypoxic culture conditions for as little as 2 days was sufficient to increase the percentage of BCSCs ≤2-fold. However, when SUM-159 breast cancer cells were stably transfected with an shRNA to block HIF-1α expression, the induction of the BCSC phenotype in response to hypoxia was abrogated, whereas expression of shRNA targeting HIF-2α had no effect [39]. This study demonstrated that HIF-1 was required for hypoxic induction of the BCSC phenotype, but did not identify any specific HIF-1 target gene products that mediated this response. Conditional knockout of HIF-1α in mouse mammary epithelial cells in MMTV-PyMT mice, which develop mammary tumours due to the expression of polyoma middle T antigen, resulted in >30-fold decrease in tumour-initiating cells [45].

Gain- and loss-of-function studies in breast cell lines suggested that another transcription factor, TAZ, also played an important role in promoting the BCSC phenotype [46]. TAZ is a co-activator that interacts with DNA-binding proteins of the TEAD (TEA/ATTS domain) family to activate the transcription of target genes, including CTGF, SERPINE1 and BIRC5, which encode connective tissue growth factor, plasminogen activator inhibitor 1 and survivin respectively [4649]. TAZ is a component of the mammalian homologue of the Hippo pathway, which was originally identified in the fruitfly Drosophila melanogaster as a critical regulator of body organ mass [50]. In mammals, phosphorylation of TAZ or its paralogue, YAP, by the kinases LATS1 and LATS2 negatively regulates its ability to enter the nucleus and bind to TEAD proteins [50]. TAZ mRNA and protein were reported to be overexpressed in ∼80% of high-grade breast cancers, and the WWTR1 gene encoding TAZ was amplified in 14% of high-grade tumours [46]. However, this left the majority of the tumours with unexplained TAZ overexpression and also did not explain how overexpressed TAZ could overcome negative regulation by LATS.

Analysis of non-metastatic MCF-7 (ER+) and HCC-1954 (HER2+), as well as metastatic MDA-MB-231 and MDA-MB-435 (TNBC), breast cancer cell lines revealed that TAZ mRNA and protein expression were increased when the cells were exposed to 1% O2 [51]. Analysis of subclones of MCF-7, MDA-MB-231 and MDA-MB-435 that were stably transfected with shRNA targeting HIF-1α or HIF-2α revealed that hypoxia-induced TAZ expression was dependent on HIF-1α. ChIP assays revealed hypoxia-induced binding of HIF-1α and HIF-1β, but not HIF-2α, to a site in intron 2 of the WWTR1 gene. Hypoxia-induced TAZ expression led to hypoxia-induced expression of the TAZ target genes CTGF, SERPINE1 and BIRC5. Knockdown of HIF-1α, but not HIF-2α, expression impaired the hypoxic induction of these genes. ChIP assays demonstrated increased binding of TAZ to TEAD-binding sites in the CTGF promoter under hypoxic conditions. Whereas TAZ was localized primarily to the cytosol of non-hypoxic cells, it was localized primarily in the nucleus of hypoxic cells. The hypoxia-induced nuclear translocation of TAZ was blocked by shRNA targeting HIF-1α [51].

These findings suggested that the negative regulation of TAZ by LATS kinases was overcome by hypoxia. Immunoblot assays of the four breast cancer cell lines revealed constitutively low levels of LATS1 expression, whereas LATS2 protein levels were high in cells cultured at 20% O2, but barely detectable in cells cultured at 1% O2. Treatment of cells with MG132 prevented the loss of LATS2 protein, indicating that hypoxia induced proteasomal degradation of LATS2. Expression of the ubiquitin protein ligase SIAH1 was induced by hypoxia and shown to ubiquitinate LATS2. Hypoxic induction of SIAH1 mRNA and protein was lost in cells with knockdown of HIF-1α, but not HIF-2α. Hypoxia-induced binding of HIF-1 to a site in intron 1 of the SIAH1 gene was demonstrated by ChIP. Knockdown of SIAH1 prevented the hypoxia-induced nuclear translocation of TAZ, whereas knockdown of LATS2 increased TAZ nuclear localization under non-hypoxic conditions [51].

Knockdown of HIF-1α or SIAH1 impaired the hypoxic induction of the BCSC phenotype as measured by ALDH+ or mammosphere-forming cells, whereas knockdown of TAZ reduced BCSCs under both hypoxic and non-hypoxic conditions and LATS2 knockdown increased BCSCs under both hypoxic and non-hypoxic conditions. Injection of 1000 control MDA-MB-231 cells into the mammary fat pad was sufficient to generate tumours in seven out of seven mice. In contrast, TAZ-knockdown MDA-MB-231 cells formed tumours in three of seven mice, and most strikingly, none of seven mice injected with HIF-1α-knockdown MDA-MB-231 cells formed tumours by 10 weeks after injection [51]. The dramatic effect of HIF-1α knockdown indicates that HIF-1 regulates additional (TAZ-independent) pathways that are important for induction of the BCSC phenotype in response to hypoxia (Figure 2). Remarkably, TAZ and HIF-1α proteins interact physically, such that TAZ is recruited to HIF-1 target genes and HIF-1α is recruited to TAZ/TEAD target genes, indicating that these two proteins serve as reciprocal co-activators with bidirectional cross-talk increasing both HIF- and TAZ-dependent transcription [52,53].

Hypoxia induces HIF-1α-dependent TAZ expression and activity

Figure 2
Hypoxia induces HIF-1α-dependent TAZ expression and activity

Intratumoral hypoxia induces HIF-1-dependent transcription of the WWTR1 gene, encoding TAZ, and the SIAH1 gene, encoding a ubiquitin protein ligase that triggers the degradation of LATS2, a kinase that phosphorylates TAZ and blocks its nuclear localization.

Figure 2
Hypoxia induces HIF-1α-dependent TAZ expression and activity

Intratumoral hypoxia induces HIF-1-dependent transcription of the WWTR1 gene, encoding TAZ, and the SIAH1 gene, encoding a ubiquitin protein ligase that triggers the degradation of LATS2, a kinase that phosphorylates TAZ and blocks its nuclear localization.

INDUCTION OF THE BCSC PHENOTYPE BY CYTOTOXIC CHEMOTHERAPY

Compared with the bulk cancer cells, BCSCs have increased resistance to chemotherapy, and the percentage of BCSCs following chemotherapy is increased compared with before therapy, a phenomenon known as BCSC enrichment [5456]. A reduction in cancer cell burden of >99% may result in an apparent complete response to therapy, but may leave behind a population of residual BCSCs that represent the source of subsequent disease recurrence and metastasis, which will eventually lead to death of the patient.

When TNBC (MDA-MB-231, SUM-149 or SUM-159) cells were exposed to cytotoxic chemotherapy (gemcitabine or paclitaxel) for 4 days at IC50 (i.e. the drug concentration that reduced cell number by 50%), both HIF-1α and HIF-2α mRNA and protein expression were significantly increased, as was the expression of HIF target genes, and HIF induction was required for BCSC enrichment [57]. The requirement for HIF-2α as well as HIF-1α for BCSC enrichment in response to chemotherapy suggested that the underlying molecular mechanism was different from induction of the BCSC phenotype in response to hypoxia.

In TNBC, IL (interleukin)-6 and IL-8 signalling pathways play major roles in stimulating BCSC survival and self-renewal [4,7,56,5864]. IL-8 is expressed by bulk tumour cells, whereas expression of its cognate receptor, CXCR1 (CXC chemokinase receptor 1), is limited to ALDH+ breast cancer cells [61]. Paclitaxel-induced IL-8 expression and BCSC enrichment were blocked by an inhibitor of TGFβ (transforming growth factor β) signalling or by siRNA against the transcription factor SMAD4 [56]. Autocrine IL-6 signalling by BCSCs induces STAT3 (signal transducer and activator of transcription 3) activity [7,62], which may further stimulate expression of IL-6 [64] and contribute to the BCSC phenotype. Paclitaxel-induced expression of IL-6 and IL-8 mRNA was partially inhibited by knockdown of HIF-1α or HIF-2α, and was completely abrogated by DKD (double knockdown) of HIF-1α and HIF-2α [57]. Thus the induction of two key cytokine signalling pathways by paclitaxel in TNBC cells is dependent on HIF activity (Figure 3). Induction of the BCSC phenotype is dependent on IL-6 and IL-8 signalling because neutralizing antibodies against either cytokine blocked chemotherapy-induced BCSC enrichment [56,57]. IL-8 induction also requires chemotherapy-induced SMAD activity [56], which is not HIF-dependent, but is not sufficient in the absence of HIF activation to induce IL-8 expression [57].

Chemotherapy induces HIF-1α- and HIF-2α-dependent expression of IL-6, IL-8 and MDR-1

Figure 3
Chemotherapy induces HIF-1α- and HIF-2α-dependent expression of IL-6, IL-8 and MDR-1

Cytotoxic chemotherapy, such as paclitaxel or gemcitabine, induces HIF-1 and HIF-2, which are required for the induction of IL-6, IL-8 and MDR1. Both IL-6 and IL-8 are required for chemotherapy-induced enrichment of BCSCs.

Figure 3
Chemotherapy induces HIF-1α- and HIF-2α-dependent expression of IL-6, IL-8 and MDR-1

Cytotoxic chemotherapy, such as paclitaxel or gemcitabine, induces HIF-1 and HIF-2, which are required for the induction of IL-6, IL-8 and MDR1. Both IL-6 and IL-8 are required for chemotherapy-induced enrichment of BCSCs.

In addition to the induction of interleukin expression, chemotherapy also induces the expression of MDR1 (multidrug-resistance protein 1), which is also known as P-glycoprotein, and paclitaxel-induced MDR1 expression was partially inhibited by knockdown of HIF-1α or HIF-2α, and completely abrogated by DKD [57]. Sorting of ALDH+ and ALDH cells revealed that, although MDR1 expression was induced in both populations of cells, it was 2-fold greater in ALDH+ than in ALDH cells, providing a direct mechanism for increased resistance of BCSCs to chemotherapy by increased drug efflux. HIF-1 has been implicated in other chemotherapy-resistance mechanisms [65], and it is possible that some of these pathways may also be induced in BCSCs.

Taken together, recent studies indicate that both intratumoral hypoxia and chemotherapy administration induce HIF activity and induce the BCSC phenotype, yet it appears that different downstream target genes may mediate the responses to hypoxia and chemotherapy, with HIF-1 mediating the BCSC response to hypoxia, whereas both HIF-1 and HIF-2 mediate the response to cytotoxic chemotherapy. It is likely that additional HIF target genes will be identified that contribute to these responses.

HIF INHIBITORS BLOCK CHEMOTHERAPY-INDUCED ENRICHMENT OF BCSCs

Drugs that inhibit HIF activity, such as digoxin, which blocks the accumulation of HIF-1α and HIF-2α protein [66], and acriflavine, which blocks the dimerization of HIF-1α or HIF-2α with HIF-1β [67], block the chemotherapy-induced enrichment of BCSCs both in vitro and in vivo [57]. Treatment of mice bearing MDA-MB-231 tumours with gemcitabine resulted in significant inhibition of growth, but, as soon as the treatment was discontinued, tumour growth immediately increased to rates that were similar to saline-treated mice; in contrast, combined treatment with gemcitabine and digoxin resulted in tumour regression and no regrowth during a 1 week follow-up after cessation of treatment [57]. Several drugs that inhibit HIF activity are currently in clinical trials, including ganetespib, which is a second-generation HSP90 (heat-shock protein 90) inhibitor that blocks the expression of multiple HIF-1 target genes, inhibits primary tumour growth, vascularization, invasion and metastasis, and, in contrast with cytotoxic chemotherapy, reduces the percentage of BCSCs in orthotopic mouse models of TNBC [68]. Another HIF inhibitor, CRLX101, reduced the number of ALDH+ cells within hypoxic regions of bevacizumab-treated tumours [69].

To explore further the clinical relevance of HIF activity in the response to chemotherapy, the expression of a HIF target gene signature (ADM, ANGPTL4, CA9, ERO1L, HIF1A, IL6, IL8, LDHA, LOX, NDRG1, P4HA2, PGK1, PLOD1, SLC2A1, SLC2A3, VEGFA mRNA) was analysed in a database of 2785 primary human breast cancers [70]. HIF signature expression greater than the median value was associated with a highly significant increase in patient mortality [HR (hazard ratio)=1.63; P=4.4×10−16]; the HR rose to 1.90 when only basal-like breast cancers were considered and reached 3.45 when only basal-like breast cancers treated with chemotherapy were analysed [57]. These analyses provide evidence that the mechanistic data obtained by analysis of human TNBC cell lines in vitro and in mice are clinically relevant.

Taken together, these studies indicate that co-administration of a HIF inhibitor may improve the clinical response to cytotoxic chemotherapy by blocking counter-therapeutic induction of the BCSC phenotype. HIF inhibitors may also block the counter-therapeutic effects of anti-angiogenesis drugs and improve their clinical efficacy by blocking hypoxia-induced enrichment of BCSCs. Although our focus has been on TNBC, the principles elucidated by these studies may have broad implications for treatment of other types of human cancer. HIFs may also play important roles in radiation resistance [71,72] and further studies are required to determine whether HIF-mediated BCSC enrichment occurs after radiation therapy.

FUNDING

G.L.S. is an American Cancer Society Research Professor and the C. Michael Armstrong Professor at the Johns Hopkins University School of Medicine. Breast cancer research in my laboratory is supported by the American Cancer Society [grant number 122437-RP-12-090-01-COUN] and the Department of Defense Breast Cancer Research Program [grant number W81XWH-12-1-0464].

Abbreviations

     
  • ALDH

    aldehyde dehydrogenase

  •  
  • BCSC

    breast cancer stem cell

  •  
  • DKD

    double knockdown

  •  
  • ER

    oestrogen receptor

  •  
  • HER2

    human epidermal growth factor receptor 2

  •  
  • HIF

    hypoxia-inducible factor

  •  
  • HR

    hazard ratio

  •  
  • IL

    interleukin

  •  
  • MDR1

    multidrug-resistance protein 1

  •  
  • PR

    progesterone receptor

  •  
  • TEAD

    TEA/ATTS domain

  •  
  • TNBC

    triple-negative breast cancer

  •  
  • VEGF

    vascular endothelial growth factor

  •  
  • VHL

    von Hippel–Lindau

References

References
1
Luo
 
M.
Clouthier
 
S.G.
Deol
 
Y.
Liu
 
S.
Azizi
 
E.
Wicha
 
M.S.
 
Breast cancer stem cells: current advances and clinical implications
Methods Mol. Biol.
2015
, vol. 
1293
 (pg. 
1
-
49
)
[PubMed]
2
Nielsen
 
T.O.
Parker
 
J.S.
Leung
 
S.
Voduc
 
D.
Ebbert
 
M.
Vickery
 
T.
Davies
 
S.R.
Snider
 
J.
Stijleman
 
I.J.
Reed
 
J.
, et al 
A comparison of PAM50 intrinsic subtyping with immunohistochemistry and clinical prognostic factors in tamoxifen-treated estrogen receptor-positive breast cancer
Clin. Cancer Res.
2010
, vol. 
16
 (pg. 
5222
-
5232
)
[PubMed]
3
Pece
 
S.
Tosoni
 
D.
Confalonieri
 
S.
Mazzarol
 
G.
Vecchi
 
M.
Ronzoni
 
S.
Bernard
 
L.
Viale
 
G.
Pelicci
 
P.G.
Di Fiore
 
P.P.
 
Biological and molecular heterogeneity of breast cancers correlates with their cancer stem cell content
Cell
2010
, vol. 
140
 (pg. 
62
-
73
)
[PubMed]
4
Ginestier
 
C.
Hur
 
M.H.
Charafe-Jauffret
 
E.
Monville
 
F.
Dutcher
 
J.
Brown
 
M.
Jacquemier
 
J.
Viens
 
P.
Kleer
 
C.G.
Liu
 
S.
, et al 
ALDH1 is a marker of normal and malignant human mammary stem cells and a predictor of poor clinical outcome
Cell Stem Cell
2007
, vol. 
1
 (pg. 
555
-
567
)
[PubMed]
5
Al-Hajj
 
M.
Wicha
 
M.S.
Benito-Hernandez
 
A.
Morrison
 
S.J.
Clarke
 
M.F.
 
Prospective identification of tumorigenic breast cancer cells
Proc. Natl. Acad. Sci. U.S.A.
2003
, vol. 
100
 (pg. 
3983
-
3988
)
[PubMed]
6
Honeth
 
G.
Bendahl
 
P.O.
Ringnér
 
M.
Saal
 
L.H.
Gruvberger-Saal
 
S.K.
Lövgren
 
K.
Grabau
 
D.
Fernö
 
M.
Borg
 
A.
Hegardt
 
C.
 
The CD44+/CD24− phenotype is enriched in basal-like breast tumors
Breast Cancer Res.
2008
, vol. 
10
 pg. 
R53
 
[PubMed]
7
Marotta
 
L.L.
Almendro
 
V.
Marusyk
 
A.
Shipitsin
 
M.
Schemme
 
J.
Walker
 
S.R.
Bloushtain-Qimron
 
N.
Kim
 
J.J.
Choudhury
 
S.A.
Maruyama
 
R.
, et al 
The JAK2/STAT3 signaling pathway is required for growth of CD44+CD24− stem cell-like breast cancer cells in human tumors
J. Clin. Invest.
2011
, vol. 
121
 (pg. 
2723
-
2735
)
[PubMed]
8
Dontu
 
G.
Abdallah
 
W.M.
Foley
 
J.M.
Jackson
 
K.W.
Clarke
 
M.F.
Kawamura
 
M.J.
Wicha
 
M.S.
 
In vitro propagation and transcriptional profiling of human mammary stem/progenitor cells
Genes Dev.
2003
, vol. 
17
 (pg. 
1253
-
1270
)
[PubMed]
9
Vaupel
 
P.
Mayer
 
A.
Höckel
 
M.
 
Tumor hypoxia and malignant progression
Methods Enzymol.
2004
, vol. 
381
 (pg. 
335
-
354
)
[PubMed]
10
Vaupel
 
P.
Höckel
 
M.
Mayer
 
A.
 
Detection and characterization of tumor hypoxia using pO2 histography
Antioxid. Redox Signal.
2007
, vol. 
9
 (pg. 
1221
-
1235
)
[PubMed]
11
Wang
 
G.L.
Semenza
 
G.L.
 
Purification and characterization of hypoxia-inducible factor 1
J. Biol. Chem.
1995
, vol. 
270
 (pg. 
1230
-
1237
)
[PubMed]
12
Wang
 
G.L.
Jiang
 
B.H.
Rue
 
E.A.
Semenza
 
G.L.
 
Hypoxia-inducible factor 1 is a basic-helix–loop–helix–PAS heterodimer regulated by cellular O2 tension
Proc. Natl. Acad. Sci. U.S.A.
1995
, vol. 
92
 (pg. 
5510
-
5514
)
[PubMed]
13
Prabhakar
 
N.R.
Semenza
 
G.L.
 
Adaptive and maladaptive cardiorespiratory responses to continuous and intermittent hypoxia mediated by hypoxia-inducible factors 1 and 2
Physiol. Rev.
2012
, vol. 
92
 (pg. 
967
-
1003
)
[PubMed]
14
Semenza
 
G.L.
 
Oxygen homeostasis
Wiley Interdiscip. Rev. Syst. Biol. Med.
2010
, vol. 
2
 (pg. 
336
-
361
)
[PubMed]
15
Zhong
 
H.
De Marzo
 
A.M.
Laughner
 
E.
Lim
 
M.
Hilton
 
D.A.
Zagzag
 
D.
Buechler
 
P.
Isaacs
 
W.B.
Semenza
 
G.L.
Simons
 
J.W.
 
Overexpression of hypoxia-inducible factor 1α in common human cancers and their metastases
Cancer Res.
1999
, vol. 
59
 (pg. 
5830
-
5835
)
[PubMed]
16
Talks
 
K.L.
Turley
 
H.
Gatter
 
K.C.
Maxwell
 
P.H.
Pugh
 
C.W.
Ratcliffe
 
P.J.
Harris
 
A.L.
 
The expression and distribution of the hypoxia-inducible factors HIF-1α and HIF-2α in normal human tissues, cancers, and tumor-associated macrophages
Am. J. Pathol.
2000
, vol. 
157
 (pg. 
411
-
421
)
[PubMed]
17
Schindl
 
M.
Schoppmann
 
S.F.
Samonigg
 
H.
Hausmaninger
 
H.
Kwasny
 
W.
Gnant
 
M.
Jakesz
 
R.
Kubista
 
E.
Birner
 
P.
Oberhuber
 
G.
 
Overexpression of hypoxia-inducible factor 1α is associated with an unfavorable prognosis in lymph node-positive breast cancer
Clin. Cancer Res.
2002
, vol. 
8
 (pg. 
1831
-
1837
)
[PubMed]
18
Bos
 
R.
van der Groep
 
P.
Greijer
 
A.E.
Shvarts
 
A.
Meijer
 
S.
Pinedo
 
H.M.
Semenza
 
G.L.
van Diest
 
P.J.
van der Wall
 
E.
 
Levels of hypoxia-inducible factor-1α independently predict prognosis in patients with lymph node negative breast carcinoma
Cancer
2003
, vol. 
97
 (pg. 
1573
-
1581
)
[PubMed]
19
Giatromanolaki
 
A.
Koukourakis
 
M.I.
Simopoulos
 
C.
Polychronidis
 
A.
Gatter
 
K.C.
Harris
 
A.L.
Sivridis
 
E.
 
c-erbB-2 related aggressiveness in breast cancer is hypoxia-inducible factor-1α dependent
Clin. Cancer Res.
2004
, vol. 
10
 (pg. 
7972
-
7977
)
[PubMed]
20
Gruber
 
G.
Greiner
 
R.H.
Hlushchuk
 
R.
Aebersold
 
D.M.
Altermatt
 
H.J.
Berclaz
 
G.
Djonov
 
V.
 
Hypoxia-inducible factor 1α in high-risk breast cancer: an independent prognostic parameter?
Breast Cancer Res.
2004
, vol. 
6
 (pg. 
R191
-
R198
)
[PubMed]
21
Dales
 
J.P.
Garcia
 
S.
Meunier-Carpentier
 
S.
Andrac-Meyer
 
L.
Haddad
 
O.
Lavaut
 
C.
Allasia
 
M.N.
Bonnier
 
P.
Charpin
 
C.
 
Overexpression of hypoxia-inducible factor HIF-1α predicts early relapse in breast cancer: retrospective study in a series of 745 patients
Int. J. Cancer
2005
, vol. 
116
 (pg. 
734
-
739
)
[PubMed]
22
Vleugel
 
M.M.
Greijer
 
A.E.
Shvarts
 
A.
van der Groep
 
P.
van Berkel
 
M.
Aarbodem
 
Y.
van Tinteren
 
H.
Harris
 
A.L.
van Diest
 
P.J.
 
van der Wall
 
E.
 
Differential prognostic impact of hypoxia-induced and diffuse HIF-1α expression in invasive breast cancer
J. Clin. Pathol.
2005
, vol. 
58
 (pg. 
172
-
177
)
[PubMed]
23
Generali
 
D.
Berruti
 
A.
Brizzi
 
M.P.
Campo
 
L.
Bonardi
 
S.
Wigfield
 
S.
Bersiga
 
A.
Allevi
 
G.
Milani
 
M.
Aguggini
 
S.
, et al 
Hypoxia-inducible factor-1α expression predicts a poor response to primary chemoendocrine therapy and disease-free survival in primary human breast cancer
Clin. Cancer Res.
2006
, vol. 
12
 (pg. 
4562
-
4568
)
[PubMed]
24
Kronblad
 
A.
Jirstrom
 
K.
Ryden
 
L.
Nordenskjold
 
B.
Landberg
 
G.
 
Hypoxia inducible factor-1α is a prognostic marker in premenopausal patients with intermediate to highly differentiated breast cancer but not a predictive marker for tamoxifen response
Int. J. Cancer
2006
, vol. 
118
 (pg. 
2609
-
2916
)
[PubMed]
25
Trastour
 
C.
Benizri
 
E.
Ettore
 
F.
Ramaioli
 
A.
Chamorey
 
E.
Pouyssegur
 
J.
Berra
 
E.
 
HIF-1α and CA IX staining in invasive breast carcinomas: Prognosis and treatment outcome
Int. J. Cancer
2007
, vol. 
120
 (pg. 
1443
-
1450
)
26
Yamamoto
 
Y.
Ibusuki
 
M.
Okumura
 
Y.
Kawasoe
 
T.
Kai
 
K.
Ivama
 
K.
Iwase
 
H.
 
Hypoxia-inducible factor 1α is closely linked to an aggressive phenotype in breast cancer
Breast Cancer Res. Treat.
2007
, vol. 
110
 (pg. 
465
-
475
)
[PubMed]
27
Passanen
 
A.
Heikkila
 
M.
Hirsila
 
M.
Kivirikko
 
K.I.
Myllyharju
 
J.
 
Hypoxia-inducible factor (HIF)-3α is subject to extensive alternative splicing in human tissues and cancer cells and is regulated by HIF-1 but not HIF-2
Int. J. Biochem. Cell Biol.
2010
, vol. 
42
 (pg. 
1189
-
1200
)
[PubMed]
28
Zhang
 
P.
Yao
 
Q.
Lu
 
L.
Li
 
Y.
Chen
 
P.J.
Duan
 
C.
 
Hypoxia-inducible factor 3 is an oxygen-dependent transcription activator and regulates a distinct transcriptional response to hypoxia
Cell Rep.
2014
, vol. 
6
 (pg. 
1110
-
1121
)
[PubMed]
29
Liao
 
D.
Corle
 
C.
Seagroves
 
T.N.
Johnson
 
R.S.
 
Hypoxia-inducible factor-1α is a key regulator of metastasis in a transgenic model of cancer initiation and progression
Cancer Res.
2007
, vol. 
67
 (pg. 
563
-
572
)
[PubMed]
30
Schito
 
L.
Rey
 
S.
Tafani
 
M.
Zhang
 
H.
Wong
 
C.C.
Russo
 
A.
Russo
 
M.A.
Semenza
 
G.L.
 
Hypoxia-inducible factor 1-dependent expression of platelet-derived growth factor B promotes lymphatic metastasis of hypoxic breast cancer cells
Proc. Natl. Acad. Sci. U.S.A.
2012
, vol. 
109
 (pg. 
E2707
-
E2716
)
[PubMed]
31
Zhang
 
H.
Wong
 
C.C.
Wei
 
H.
Gilkes
 
D.M.
Korangath
 
P.
Chaturvedi
 
P.
Schito
 
L.
Chen
 
J.
Krishnamachary
 
B.
Winnard
 
P.T.
, et al 
HIF-1-dependent expression of angiopoietin-like 4 and L1CAM mediates vascular metastasis of hypoxic breast cancer cells to the lungs
Oncogene
2012
, vol. 
31
 (pg. 
1757
-
1770
)
[PubMed]
32
Doedens
 
A.L.
Stockmann
 
C.
Rubinstein
 
M.P.
Liao
 
D.
Zhang
 
N.
DeNardo
 
D.G.
Coussens
 
L.M.
Karin
 
M.
Goldrath
 
A.W.
Johnson
 
R.S.
 
Macrophage expression of hypoxia-inducible factor-1α suppresses T-cell function and promotes tumor progression
Cancer Res.
2010
, vol. 
70
 (pg. 
7465
-
7475
)
[PubMed]
33
Branco-Price
 
C.
Zhang
 
N.
Schnelle
 
M.
Evans
 
C.
Katschinski
 
D.M.
Liao
 
D.
Ellies
 
L.
Johnson
 
R.S.
 
Endothelial cell HIF-1α and HIF-2α differentially regulate metastatic success
Cancer Cell
2012
, vol. 
21
 (pg. 
52
-
65
)
[PubMed]
34
Kim
 
J.W.
Evans
 
C.
Weidemann
 
A.
Takeda
 
N.
Lee
 
Y.S.
Stockmann
 
C.
Branco-Price
 
C.
Brandberg
 
F.
Leone
 
G.
Ostrowski
 
M.C.
, et al 
Loss of fibroblast HIF-1α accelerates tumorigenesis
Cancer Res.
2012
, vol. 
72
 (pg. 
3187
-
3195
)
[PubMed]
35
Wang
 
T.
Gilkes
 
D.M.
Takano
 
N.
Xiang
 
L.
Luo
 
W.
Bishop
 
C.J.
Chaturvedi
 
P.
Green
 
J.J.
Semenza
 
G.L.
 
Hypoxia-inducible factors and RAB22A mediate formation of microvesicles that stimulate breast cancer invasion and metastasis
Proc. Natl. Acad. Sci. U.S.A.
2014
, vol. 
111
 (pg. 
E3234
-
E3242
)
[PubMed]
36
Luo
 
W.
Chang
 
R.
Zhong
 
J.
Pandey
 
A.
Semenza
 
G.L.
 
Histone demethylase JMJD2C is a coactivator for hypoxia-inducible factor 1 that is required for breast cancer progression
Proc. Natl. Acad. Sci. U.S.A.
2012
, vol. 
109
 (pg. 
E3367
-
E3376
)
[PubMed]
37
Oskarsson
 
T.
Batlle
 
E.
Massague
 
J.
 
Metastatic stem cells: sources, niches, and vital pathways
Cell Stem Cell
2014
, vol. 
14
 (pg. 
306
-
321
)
[PubMed]
38
Mathieu
 
J.
Zhang
 
Z.
Zhou
 
W.
Wang
 
A.J.
Heddleston
 
J.M.
Pinna
 
C.M.
Hubaud
 
A.
Stadler
 
B.
Choi
 
M.
Bar
 
M.
, et al 
HIF induces human embryonic stem cell markers in cancer cells
Cancer Res.
2011
, vol. 
71
 (pg. 
4640
-
4652
)
[PubMed]
39
Conley
 
S.J.
Gheordunescu
 
E.
Kakarala
 
P.
Newman
 
B.
Korkaya
 
H.
Heath
 
A.N.
Clouthier
 
S.G.
Wicha
 
M.S.
 
Antiangiogenic agents increase breast cancer stem cells via the generation of tumor hypoxia
Proc. Natl. Acad. Sci. U.S.A.
2012
, vol. 
109
 (pg. 
2784
-
2789
)
[PubMed]
40
Forsythe
 
J.A.
Jiang
 
B.H.
Iyer
 
N.V.
Agani
 
F.
Leung
 
S.W.
Koos
 
R.D.
Semenza
 
G.L.
 
Activation of vascular endothelial growth factor gene transcription by hypoxia-inducible factor 1
Mol. Cell. Biol.
1996
, vol. 
16
 (pg. 
4604
-
4613
)
[PubMed]
41
Ferrara
 
N.
 
VEGF as a therapeutic target in cancer
Oncology
2005
, vol. 
69
 (pg. 
11
-
16
)
[PubMed]
42
Rose
 
S.
 
FDA pulls approval for avastin in breast cancer
Cancer Discov.
2011
, vol. 
1
 (pg. 
OF1
-
OF2
)
43
Ebos
 
J.M.
Lee
 
C.R.
Cruz-Munoz
 
W.
Bjarnason
 
G.A.
Christensen
 
J.G.
Kerbel
 
R.S.
 
Accelerated metastasis after short-term treatment with a potent inhibitor of tumor angiogenesis
Cancer Cell
2009
, vol. 
15
 (pg. 
232
-
239
)
[PubMed]
44
Pàez-Ribes
 
M.
Allen
 
E.
Hudock
 
J.
Takeda
 
T.
Okuyama
 
H.
Viñals
 
F.
Inoue
 
M.
Bergers
 
G.
Casanovas
 
O
 
Antiangiogenic therapy elicits malignant progression of tumors to increased local invasion and distant metastasis
Cancer Cell
, vol. 
3
 (pg. 
220
-
231
)
45
Schwab
 
L.P.
Peacock
 
D.L.
Majumdar
 
D.
Ingels
 
J.F.
Jensen
 
L.C.
Smith
 
K.D.
Cushing
 
R.C.
Seagroves
 
T.N.
 
Hypoxia-inducible factor 1α promotes primary tumor growth and tumor-initiating cell activity in breast cancer
Breast Cancer Res.
2012
, vol. 
14
 pg. 
R6
 
[PubMed]
46
Cordenonsi
 
M.
Zanconato
 
F.
Azzolin
 
L.
Forcato
 
M.
Rosato
 
A.
Frasson
 
C.
Inui
 
M.
Montagner
 
M.
Parenti
 
A.R.
Poletti
 
A.
, et al 
The Hippo transducer TAZ confers cancer stem cell-related traits on breast cancer cells
Cell
2011
, vol. 
147
 (pg. 
759
-
772
)
[PubMed]
47
Zhang
 
H.
Liu
 
C.Y.
Zha
 
Z.Y.
Zhao
 
B.
Yao
 
J.
Zhao
 
S.
Xiong
 
Y.
Lei
 
Q.Y.
Guan
 
K.L.
 
TEAD transcription factors mediate the function of TAZ in cell growth and epithelial–mesenchymal transition
J. Biol. Chem.
2009
, vol. 
284
 (pg. 
13355
-
13362
)
[PubMed]
48
Lai
 
D.
Ho
 
K.C.
Hao
 
Y.
Yang
 
X.
 
Taxol resistance in breast cancer cells is mediated by the hippo pathway component TAZ and its downstream transcriptional targets Cyr61 and CTGF
Cancer Res.
2011
, vol. 
71
 (pg. 
2728
-
2738
)
[PubMed]
49
Pobbati
 
A.V.
Hong
 
W.
 
Emerging roles of TEAD transcription factors and its coactivators in cancers
Cancer Biol. Ther.
2013
, vol. 
14
 (pg. 
390
-
398
)
[PubMed]
50
Pan
 
D.
 
The hippo signaling pathway in development and cancer
Dev. Cell
2010
, vol. 
19
 (pg. 
491
-
505
)
[PubMed]
51
Xiang
 
L.
Gilkes
 
D.M.
Hu
 
H.
Takano
 
N.
Luo
 
W.
Lu
 
H.
Bullen
 
J.W.
Samanta
 
D.
Liang
 
H.
Semenza
 
G.L.
 
Hypoxia-inducible factor 1 mediates TAZ expression and nuclear localization to induce the breast cancer stem cell phenotype
Oncotarget
2014
, vol. 
5
 (pg. 
12509
-
12527
)
[PubMed]
52
Bendinelli
 
P.
Maroni
 
P.
Matteucci
 
E.
Luzzati
 
A.
Perrucchini
 
G.
Desiderio
 
M.A.
 
Hypoxia inducible factor-1 is activated by transcriptional co-activator with PDZ-binding motif (TAZ) versus WWdomain-containing oxidoreductase (WWOX) in hypoxic microenvironment of bone metastasis from breast cancer
Eur. J. Cancer
2013
, vol. 
49
 (pg. 
2608
-
2618
)
[PubMed]
53
Xiang
 
L.
Gilkes
 
D.M.
Hu
 
H.
Luo
 
W.
Bullen
 
J.W.
Liang
 
H.
Semenza
 
G.L.
 
HIF-1α and TAZ serve as reciprocal co-activators in human breast cancer cells
Oncotarget
2015
, vol. 
6
 (pg. 
11768
-
11778
)
[PubMed]
54
Li
 
X.
Lewis
 
M.T.
Huang
 
J.
Gutierrez
 
C.K.
Osborne
 
C.K.
Wu
 
M.F.
Hilsenbeck
 
S.G.
Pavlick
 
A.
Zhang
 
X.
Chamness
 
G.C.
, et al 
Intrinsic resistance of tumorigenic breast cancer cells to chemotherapy
J. Natl. Cancer Inst.
2008
, vol. 
100
 (pg. 
672
-
679
)
[PubMed]
55
Creighton
 
C.J.
Li
 
X.
Landis
 
M.
Dixon
 
J.M.
Neumeister
 
V.M.
Sjolund
 
A.
Rimm
 
D.L.
Wong
 
H.
Rodriguez
 
A.
Herschkowitz
 
J.I.
, et al 
Residual breast cancers after conventional therapy display mesenchymal as well as tumor-initiating features
Proc. Natl. Acad. Sci. U.S.A.
2009
, vol. 
106
 (pg. 
13820
-
13825
)
[PubMed]
56
Bhola
 
N.E.
Balko
 
J.M.
Dugger
 
T.C.
Kuba
 
M.G.
Sánchez
 
V.
Sanders
 
M.
Stanford
 
J.
Cook
 
R.S.
Arteaga
 
C.L.
 
TGF-β inhibition enhances chemotherapy action against triple-negative breast cancer
J. Clin. Invest.
2013
, vol. 
123
 (pg. 
1348
-
1358
)
[PubMed]
57
Samanta
 
D.
Gilkes
 
D.M.
Chaturvedi
 
P.
Xiang
 
L.
Semenza
 
G.L.
 
Hypoxia-inducible factors are required for chemotherapy resistance of breast cancer stem cells
Proc. Natl. Acad. Sci. U.S.A.
2014
, vol. 
111
 (pg. 
E5429
-
E5438
)
[PubMed]
58
Ponti
 
D.
Costa
 
A.
Zaffaroni
 
N.
Pratesi
 
G.
Petrangolini
 
G.
Coradini
 
D.
Pilotti
 
S.
Pierotti
 
M.A.
Daidone
 
M.G.
 
Isolation and in vitro propagation of tumorigenic breast cancer cells with stem/progenitor cell properties
Cancer Res.
2005
, vol. 
65
 (pg. 
5506
-
5511
)
[PubMed]
59
Sansone
 
P.
Storci
 
G.
Tavolari
 
S.
Guarnieri
 
T.
Giovannini
 
C.
Taffurelli
 
M.
Ceccarelli
 
C.
Santini
 
D.
Paterini
 
P.
Marcu
 
K.B.
, et al 
IL-6 triggers malignant features in mammospheres from human ductal breast carcinoma and normal mammary gland
J. Clin. Invest.
2007
, vol. 
117
 (pg. 
3988
-
4002
)
[PubMed]
60
Charafe-Jauffret
 
E.
Ginestier
 
C.
Iovino
 
F.
Wicinski
 
J.
Cervera
 
N.
Finetti
 
P.
Hur
 
M.H.
Diebel
 
M.E.
Monville
 
F.
Dutcher
 
J.
, et al 
Breast cancer cell lines contain functional cancer stem cells with metastatic capacity and a distinct molecular signature
Cancer Res.
2009
, vol. 
69
 (pg. 
1302
-
1313
)
[PubMed]
61
Ginestier
 
C.
Liu
 
S.
Diebel
 
M.E.
Korkaya
 
H.
Luo
 
M.
Brown
 
M.
Wicinski
 
J.
Cabaud
 
O.
Charafe-Jauffret
 
E.
Birnbaum
 
D.
, et al 
CXCR1 blockade selectively targets human breast cancer stem cells in vitro and in xenografts
J. Clin. Invest.
2010
, vol. 
120
 (pg. 
485
-
497
)
[PubMed]
62
Hartman
 
Z.C.
Poage
 
G.M.
den Hollander
 
P.
Tsimelzon
 
A.
Hill
 
J.
Panupinthu
 
N.
Zhang
 
Y.
Mazumdar
 
A.
Hilsenbeck
 
S.G.
Mills
 
G.B.
, et al 
Growth of triple-negative breast cancer cells relies upon coordinate autocrine expression of the proinflammatory cytokines IL-6 and IL-8
Cancer Res.
2011
, vol. 
73
 (pg. 
3470
-
3480
)
63
Iliopoulos
 
D.
Hirsch
 
H.A.
Wang
 
G.
Struhl
 
K.
 
Inducible formation of breast cancer stem cells and their dynamic equilibrium with non-stem cancer cells via IL-6 secretion
Proc. Natl. Acad. Sci. U.S.A.
2011
, vol. 
108
 (pg. 
1397
-
1402
)
[PubMed]
64
Przanowski
 
P.
Dabrowski
 
M.
Ellert-Miklaszewska
 
A.
Kloss
 
M.
Mieczkowski
 
J.
Kaza
 
B.
Ronowicz
 
A.
Hu
 
F.
Piotrowski
 
A.
Kettenmann
 
H.
, et al 
The signal transducers Stat1 and Stat3 and their novel target Jmjd3 drive the expression of inflammatory genes in microglia
J. Mol. Med.
2014
, vol. 
92
 (pg. 
239
-
254
)
[PubMed]
65
Rohwer
 
N.
Cramer
 
T.
 
Hypoxia-mediated drug resistance: novel insights on the functional interaction of HIFs and cell death pathways
Drug Resist. Updat.
2011
, vol. 
14
 (pg. 
191
-
201
)
[PubMed]
66
Zhang
 
H.
Qian
 
D.Z.
Tan
 
Y.S.
Lee
 
K.
Gao
 
P.
Ren
 
Y.R.
Rey
 
S.
Hammers
 
H.
Chang
 
D.
Pili
 
R.
, et al 
Digoxin and other cardiac glycosides inhibit HIF-1α synthesis and block tumor growth
Proc. Natl. Acad. Sci. U.S.A.
2008
, vol. 
105
 (pg. 
19579
-
19586
)
[PubMed]
67
Lee
 
K.
Zhang
 
H.
Qian
 
D.Z.
Rey
 
S.
Liu
 
J.O.
Semenza
 
G.L.
 
Acriflavine inhibits HIF-1 dimerization, tumor growth, and vascularization
Proc. Natl. Acad. Sci. U.S.A.
2009
, vol. 
106
 (pg. 
17910
-
17915
)
[PubMed]
68
Xiang
 
L.
Gilkes
 
D.M.
Chaturvedi
 
P.
Luo
 
W.
Hu
 
H.
Takano
 
N.
Liang
 
H.
Semenza
 
G.L.
 
Ganetespib blocks HIF-1 activity and inhibits tumor growth, vascularization, stem cell maintenance, invasion, and metastasis in orthotopic mouse models of triple-negative breast cancer
J. Mol. Med.
2014
, vol. 
92
 (pg. 
151
-
164
)
[PubMed]
69
Conley
 
S.J.
Baker
 
T.L.
Burnett
 
J.P.
Theisen
 
R.L.
Lazarus
 
D.
Peters
 
C.G.
Clouthier
 
S.G.
Eliasof
 
S.
Wicha
 
M.S.
 
CRLX101, an investigational camptothecin-containing nanoparticle-drug conjugate, targets cancer stem cells and impedes resistance to antiangiogenic therapy in mouse models of breast cancer
Breast Cancer Res. Treat.
2015
, vol. 
150
 (pg. 
559
-
567
)
[PubMed]
70
Győrffy
 
B.
Lánczky
 
A.
Eklund
 
A.C.
Denkert
 
C.
Budczies
 
J.
Li
 
Q.
Szallasi
 
Z.
 
An online survival analysis tool to rapidly assess the effect of 22,277 genes on breast cancer prognosis using microarray data of 1,809 patients
Breast Cancer Res. Treat.
2010
, vol. 
123
 (pg. 
725
-
731
)
[PubMed]
71
Moeller
 
B.
Dewhirst
 
M.W.
 
HIF-1 and tumor radiosensitivity
Br. J. Cancer
2006
, vol. 
95
 (pg. 
1
-
5
)
[PubMed]
72
Zhang
 
M.
Qiu
 
Q.
Li
 
Z.
Sachdeva
 
M.
Min
 
H.
Cardona
 
D.M.
DeLaney
 
T.F.
Han
 
T.
Ma
 
Y.
Luo
 
L.
, et al 
HIF-1α regulates the response of primary sarcomas to radiation therapy through a cell autonomous mechanism
Radiat. Res.
2015
, vol. 
183
 (pg. 
594
-
609
)
[PubMed]
73
Chaturvedi
 
P.
Gilkes
 
D.M.
Wong
 
C.C.
Kshitiz, Luo
 
W.
Zhang
 
H.
Wei
 
H.
Takano
 
N.
Schito
 
L.
Levchenko
 
A.
, et al 
Hypoxia-inducible factor-dependent breast cancer–mesenchymal stem cell bidirectional signaling promotes metastasis
J. Clin. Invest.
2013
, vol. 
123
 (pg. 
189
-
205
)
[PubMed]
74
Chaturvedi
 
P.
Gilkes
 
D.M.
Takano
 
N.
Semenza
 
G.L.
 
Hypoxia-inducible factor-dependent signaling between triple-negative breast cancer cells and mesenchymal stem cells promotes macrophage recruitment
Proc. Natl. Acad. Sci. U.S.A.
2014
, vol. 
111
 (pg. 
E2120
-
E2129
)
[PubMed]
75
Gilkes
 
D.M.
Xiang
 
L.
Lee
 
S.J.
Chaturvedi
 
P.
Hubbi
 
M.E.
Wirtz
 
D.
 
Semenza
 
G.L.
 
Hypoxia-inducible factors mediate coordinated RhoA–ROCK1 expression and signaling in breast cancer cells
Proc. Natl. Acad. Sci. U.S.A.
2014
, vol. 
111
 (pg. 
E384
-
E393
)
[PubMed]
76
Gilkes
 
D.M.
Bajpai
 
S.
Wong
 
C.C.
Chaturvedi
 
P.
Hubbi
 
M.E.
Wirtz
 
D.
Semenza
 
G.L.
 
Procollagen lysyl hydroxylase 2 is essential for hypoxia-induced breast cancer metastasis
Mol. Cancer Res.
2013
, vol. 
11
 (pg. 
456
-
466
)
[PubMed]
77
Gilkes
 
D.M.
Chaturvedi
 
P.
Bajpai
 
S.
Wong
 
C.C.
Wei
 
H.
Pitcairn
 
S.
Hubbi
 
M.E.
Wirtz
 
D.
Semenza
 
G.L.
 
Collagen prolyl hydroxylases are essential for breast cancer metastasis
Cancer Res.
2013
, vol. 
73
 (pg. 
3285
-
3296
)
[PubMed]
78
Erler
 
J.T.
Bennewith
 
K.L.
Nicolau
 
M.
Dornhöfer
 
N.
Kong
 
C.
Le
 
Q.T.
Chi
 
J.T.
Jeffrey
 
S.S.
Giaccia
 
A.J.
 
Lysyl oxidase is essential for hypoxia-induced metastasis
Nature
2006
, vol. 
440
 (pg. 
1222
-
1226
)
[PubMed]
79
Erler
 
J.T.
Bennewith
 
K.L.
Cox
 
T.R.
Lang
 
G.
Bird
 
D.
Koong
 
A.
Le
 
Q.T.
Giaccia
 
A.J.
 
Hypoxia-induced lysyl oxidase is a critical mediator of bone marrow cell recruitment to form the premetastatic niche
Cancer Cell
2009
, vol. 
15
 (pg. 
35
-
44
)
[PubMed]
80
Wong
 
C.C.
Gilkes
 
D.M.
Zhang
 
H.
Chen
 
J.
Wei
 
H.
Chaturvedi
 
P.
Fraley
 
S.I.
Wong
 
C.M.
Khoo
 
U.S.
Ng
 
I.O.
, et al 
Hypoxia-inducible factor 1 is a master regulator of breast cancer metastatic niche formation
Proc. Natl. Acad. Sci. U.S.A.
2011
, vol. 
108
 (pg. 
16369
-
16374
)
[PubMed]