Generation of Breast Cancer Stem Cells through
Anne-Pierre Morel1, Marjory Lie `vre1, Cle ´mence Thomas1,2, George Hinkal1,2, Ste ´phane Ansieau1,2, Alain
1Centre Le ´on Be ´rard, Lyon, France, 2Inserm, U590, Lyon, France, 3Universite ´ de Lyon, Lyon1, ISPB, Lyon, France
Recently, two novel concepts have emerged in cancer biology: the role of so-called ‘‘cancer stem cells’’ in tumor initiation,
and the involvement of an epithelial-mesenchymal transition (EMT) in the metastatic dissemination of epithelial cancer cells.
Using a mammary tumor progression model, we show that cells possessing both stem and tumorigenic characteristics of
‘‘cancer stem cells’’ can be derived from human mammary epithelial cells following the activation of the Ras-MAPK pathway.
The acquisition of these stem and tumorigenic characters is driven by EMT induction.
Citation: Morel A-P, Lie `vre M, Thomas C, Hinkal G, Ansieau S, et al. (2008) Generation of Breast Cancer Stem Cells through Epithelial-Mesenchymal Transition. PLoS
ONE 3(8): e2888. doi:10.1371/journal.pone.0002888
Editor: Juha Klefstrom, University of Helsinki, Finland
Received March 14, 2008; Accepted July 14, 2008; Published August 6, 2008
Copyright: ? 2008 Morel et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This work was granted by the Institut National du Cancer, the Ligue contre le Cancer and the Breast Cancer Research Foundation.
Competing Interests: The authors have declared that no competing interests exist.
* E-mail: firstname.lastname@example.org
A growing body of evidence supports the notion that only a
small subset of cells within a tumor, termed cancer stem cells
(CSCs) or tumor-initiating cells, are capable of both tumor
initiation and sustaining tumor growth . Two basic arguments
underlie the hypothesis that cancer stem cells originate from
normal tissue stem cells. First, as tumor development is believed to
result from the sequential and progressive accumulation of genetic
abnormalities, adult stem cells appear to be ideal initial targets for
malignant transformation due to their long lifespans. Second,
CSCs share several properties with normal stem cells, such as their
capacity for self-renewal and their ability to differentiate [2,3].
The notion of a stem cell origin of cancer was first introduced in
the context of hematological malignancies. This hypothesis has
been supported by accumulating evidence in both chronic and
acute leukemias [4–6]. Additionally, committed hematopoietic
progenitor cells, with no inherent self-renewal properties, can be
induced to generate cells capable of initiating and maintaining
leukemias using leukemogenic fusion proteins [7–10], indicating
that there is no absolute prerequisite for genetic mutation of
normal stem cells.
Over the past few years, candidate cancer stem cells have been
identified in a variety of human malignancies including leukemias
and a number of solid tumors such as glioblastomas, medulloblas-
tomas and carcinomas [11–24]. Breast cancer is the first human
carcinoma for which a putative cancer stem cell subpopulation has
been isolated . Using in vitro-separated tumorigenic cells from
malignant human breast cancer-derived pleural effusions, Al Hajj
expression and low or undetectable levels of CD24 (CD44+CD242/
low) . These cells were highly tumorigenic when injected into
immunocompromised NOD/SCID mice and shared classic features
of normal stem cells, including the capacity for self-renewal and
generation of heterogeneous progeny . The stem/progenitor cell
phenotype of these cells was further refined by the Daidone group,
who were able to grow mammospheres from single-cell suspensions
obtained from the dissociation of primary breast tumors .
Mammospheres are non-adherent spherical cell clusters obtained
in selective culture conditions, that have been shown to be enriched
in mammary stem/progenitor cells . The vast majority of cells in
culture were CD44+CD242/low, and 10 to 20% of these retained the
ability to self-renew .
Congruent with previously reported experiments using models of
hematopoietic malignancies [7–10], transformed breast cancer cells
were obtained in vitro by introducing a series of oncogenes and
cancer-associated genes into normal primary human mammary
epithelial cells. This experimental system starts with primary human
mammary epithelial cells (HMECs), that undergo sequential
retroviral-mediated expression of the telomerase catalytic subunit
(giving rise to HMEC/hTERT cells), SV40 large T and small t
antigens (HMLE cells) and an oncogenic allele of H-Ras, H-RasV12
(HMLER cells) . Using this model, we demonstrate that
CD44+CD242/lowcells possessing stem-like properties can be
generated from CD44lowCD24+non-tumorigenic mammary epi-
thelial cells through activation of the Ras/MAPK signaling pathway
and can be accelerated by EMT induction.
CD44+CD242/lowcells, we implemented a model of human
breast cancer progression described by Elenbaas et al. 
(Figure 1A). The Weinberg’s group showed that HMLER cells
(HMECs overexpressing hTERT, SV40 T/t and H-RasV12) were
tumorigenic when injected subcutaneously or into the mammary
glands of immunocompromised mice, suggesting the possible
generation of cancer stem cells. To test this hypothesis, we
analyzed the capabilities of the different cell populations (HMECs,
HMEC/hTERT, HMLE and HMLER) to grow as non-adherent
PLoS ONE | www.plosone.org1August 2008 | Volume 3 | Issue 8 | e2888
mammospheres, a property associated with mammary stem/
progenitor cells [3,26]. In contrast with HMECs, HMEC/hTERT
and HMLE cells (Figure 1B–D), only HMLER cells grew as non-
adherent clusters, suggesting that they displayed both tumorigenic
and stem-like properties (Figure 1E). We further analyzed the
phenotype of the different HMEC-derived cell lines (expanded in
adherent conditions) by fluorescence-activated cell sorting (FACS)
using CD44 and CD24 as markers (Figure 1F–I). HMECs and
HMEC/hTERT immortalized mammary epithelial cells were
CD44+CD242/lowcells (subsequently referred to as CD242) were
totally undetectable in both cell populations (Figures 1F and 1G).
The additional introduction of oncogenes was associated with the
appearance and progressive increase of CD242cells which
accounted for 1.4% of HMLE cells and more than 65% of
HMLER cells (Figures 1H and 1I). Consistent with the report that
several components of the Ras/MAPK pathway are present in the
expression profile of CD242cancer cells , this observation
suggested that cancer stem cells could be generated in response to
the activation of specific signal transduction pathways.
This hypothesis was further investigated by studying using
FACS analysis the emergence of CD242cells following retroviral
expression of H-RasV12in HMLE cells. Whereas this cell
population remained low (,2%) in uninfected cells, mutant Ras
expression caused its progressive accumulation from 3.2% at day 5
following infection, 10.1% at day 24, 32.1% at day 30, to 65.4% at
day 55 (Figure 1I and data not shown). This observation cannot be
interpreted as a consequence of an enrichment of a rare
subpopulation of cells displaying growth advantage because
CD242and CD24+cells showed similar proliferation potential,
as demonstrated by the methylthiazolyldiphenyl-tetrazolium
bromide (MTT) assay (data not shown). Of note, the length of
time necessary for the emergence of CD242cells following H-
RasV12introduction is consistent with the initial observation by
referredto as CD24+) and
Elenbaas and colleagues that the tumorigenicity of HMLER cells
was associated with the occurrence of secondary events . In
order to determine the origin of CD242cells, we performed cell
sorting experiments and single-cell cloning assays following
retroviral expression of H-RasV12in HMLE cells. After sorting
of CD242and CD24+populations, CD24+cells were seeded onto
96-well plates under limiting dilution cloning conditions in order
to isolate single cells (day 30 post-infection). After three weeks of
cell growth, 35 independent individual clones were isolated and
characterized. Whereas 47% of the clones were fully composed of
CD24+cells, 33% displayed a heterogeneous population of CD24+
and CD242cells, and 19% were homogeneous for CD242cells.
These observations demonstrate that CD242cells can originate
from CD24+cells, since a single CD24+cell is able to generate
either heterogeneous CD24+/CD242clones or homogeneous
We next evaluated the transformation and stem-like properties
of CD24+and CD242cells generated by retroviral expression of
H-RasV12in HMLE cells. Unlike CD24+cells, CD242cells were
able to grow in soft agar, a characteristic of transformed cells
(Figure 2A). The CD242population was also able to form tumors
when injected into the mammary fat pads of nude mice (6/6 for
CD242versus 0/6 for CD24+cells, Figure 2B). This difference was
shown to be independent of the level of expressed Ras (data not
shown). Additionally, in non-adherent conditions, only CD242
cells generated mammospheres, further demonstrating the stem-
like properties of HMLER cells (Figure 2C). Of note, similar
results were invariably obtained using either the sorted CD242cell
population or CD242isolated clones (data not shown). Primary
spheres obtained from CD242cells could be enzymatically
dissociated with trypsin into single cells to give rise to secondary
spheres (data not shown).
We next attempted to validate our observations in a different
cellular context using the immortal human mammary epithelial
Figure 1. Characterization of the different steps of the in vitro model of HMEC transformation. (A) Schematic representation of the
successive steps of the transformation. (B,C,D,E) Evaluation of the ability of HMEC-derived cell lines to grow as non-adherent mammospheres. Scale
bars=100 mm. FACS analysis of CD24 and CD44 markers in HMEC-derived cell lines (F,G,H,I). B and F: Primary Human Mammary Epithelial Cell,
(HMECs); C and G: hTERT- immortalized HMECs (HME); D and H: HME expressing SV40 small t and large T antigens (HMLE); E and I: H-RasV12-infected
HMLE cells (HMLER).
Cancer Stem Cells and EMT
PLoS ONE | www.plosone.org2August 2008 | Volume 3 | Issue 8 | e2888
cell line MCF10A. As we demonstrated in transformed HMECs,
MCF10A infection with K-RasV12, was associated with the
emergence of CD242cells: 1% in MCF10A cells infected with
empty retroviral vector as compared to 90% in those infected with
the K-RasV12expressing construct (Figure 3A). Intriguingly,
enrichment in CD242cells in both HMLER and MCF10A cell
lines was coincided with a morphological change in the cells from a
general epithelial appearance to a dominant population of spindle
cells, characteristic of mesenchymal cells (Figure 3B). To examine
whether CD242cells might emerge through an EMT process, a
morphogenetic process in which cells lose their epithelial character-
istics and gain mesenchymal properties, the expression levels of
epithelial and mesenchymal markers in both CD24+and CD242
cells, was assessed. In contrast to CD24+cells, CD242cells, either
sorted from the bulk population or cloned, expressed low or
undetectable levels of epithelial markers (E-cadherin and b-catenin)
and high levels of mesenchymal markers (vimentin and fibronectin),
suggesting that they underwent an EMT (Figure 3C). To evaluate
such a possibility, we thus examined whether the treatment of
HMLER CD24+with TGFb1, a potent inducer of EMT, led to
CD242cell appearance. Indeed, eight days after treatment, a
concomitant enrichment of mesenchymal cells (characterized by a
loss of the E-Cadherinand the induction ofvimentin expression) and
As Ras and TGFb are known to cooperate in EMT induction, we
then assumed that, in presence of TGFb, the length of time needed
to generate CD242cells from CD24+HMLE cells following
infection with Ras should be shortened. As shown in Figure 5, in
agreement with our hypothesis, the addition of TGFb1 significantly
accelerated the emergence of CD242cells from CD24+HMLE cells
infected with H-RasV12.
CD44+CD242/low(CD242) cells can originate from primary
(HMECs) following their transformation with a limited number
of oncogenes and cancer-associated genes. Specifically, activation of
the Ras signaling pathway appears to be a crucial event to facilitate
the emergence of CD242cells. Strikingly, in both HMECs and
MCF10A cells, the CD242phenotype was constantly associated
with features of an epithelial-mesenchymal transition (EMT),
including the loss of epithelial markers and the concomitant gain
of mesenchymal markers. We then assumed that CD242cells could
arise from CD24+through an EMT trans-differentiation process.
Accordingly, we showed a cooperative effect of TGFb and Ras
activation as treatment of Ras infected cells with TGFb1 accelerates
recent observation that CD242cells isolated from breast cancer
tissues display a mesenchymal phenotype attributable to the
activation of TGFb and Wnt signaling , two pathways known
to be involved in EMT .
EMT, which was first recognized as a crucial feature of
embryogenesis, converts epithelial cells into mesenchymal cells
through profound disruption of cell-cell junctions and extensive
reorganization of the actin cytoskeleton . Although still
controversial, this process is presumed to be required for tumor
invasion and metastasis of carcinoma cells by promoting loss of
contact inhibition, increased cell motility and enhanced invasive-
ness . EMT is believed to be governed by complex networks
largely influenced by signals from the neoplastic microenviron-
ment. Indeed, in vitro, a variety of cytokines, including TGFb and
growth factors like hepatocyte growth factor (HGF), epidermal
growth factor (EGF) or fibroblast growth factors (FGFs), can
trigger EMT after activation of their cognate receptors in specific
cell types. Of note, growth factors transduce signals through the
activation of their cognate receptor tyrosine kinases and of their
central downstream effector Ras, giving a rationale for the
cooperative effect of Ras and TGFb in EMT promotion [34,35].
In the experimental model of breast cancer progression, the
introduction of an activated version of Ras constitutes the initial
event that sensitizes mammary epithelial cells to EMT. However,
the delay for EMT induction and the associated emergence of H-
RasV12-transformed CD242cells suggests that additional events
are required. These additional events may depend upon
Figure 2. CD242cells display tumorigenic and stem-like properties. (A) Colony assay. Growth in soft agar. Number of colonies are indicated
for 56103plated cells (+/2 standard deviation, n=3, 40x magnification). (B) Tumorigenicity assay. Athymic nude mice received a single injection of
106CD24+or CD242cells to a mammary fat pad. Tumor growth was monitored every 3 days and volume measured every 15 days. Blue lines indicate
CD242cells and red lines indicate CD24+cells, marker shapes represent duplicate injections in separate animals from three independent clones for
both cell populations. (C) In contrast to CD24+cells, CD242cells formed mammospheres in low-adherent conditions. Scale bars=100 mm.
Cancer Stem Cells and EMT
PLoS ONE | www.plosone.org3August 2008 | Volume 3 | Issue 8 | e2888
environmental EMT-inducing signals since the addition of TGFb
to the culture medium significantly decreases the time required for
completing the process and increases the percentage of CD242
cells. Altogether, our observations support the intriguing hypoth-
esis that the CD44+CD242/lowcells (or at least a fraction of them)
present within a primary breast cancer might reflect the propensity
of malignant cells to undergo transdifferentiation and metastasize.
Considering the role of EMT in invasiveness and metastatic
dissemination, our observations provide a rational explanation for
the prognostic value of the gene-expression signature of CD242
cancer stem cells in breast cancers . This signature, termed IGS
(for invasiveness gene signature), has been generated by the Clarke
group by comparing the gene-expression profiles of CD242breast-
cancer cells and normal breast epithelium. Importantly, the IGS is
significantly associated with both overall survival and metastasis-free
survival in patients with breast cancer or with other types of
malignancy . Nevertheless, this observation is at odds with the
description of cancer stem cells as a minority population within a
tumor, because the gene expression profile of these rare cells is likely
to be masked when tumors as a whole are analyzed for gene
expression . Our findings strongly suggest that the IGS might be
a consequence of the oncogenic activation of signaling pathways
involved in EMT, invasion and metastasis. This hypothesis is further
substantiated by the presence of components of the Ras/MAPK
pathway in the IGS, as well as targets of TGFb, and inducers of
EMT and/or mesenchymal markers (MGP, CXCL12, MMP-7,
Ets1, Ezrin, Wee1) . It is also highly consistent with the recent
observation thatbreastcancer cell linescontaininga highpercentage
of CD242cells express basal/mesenchymal markers and display
invasive properties .
Figure 3. Oncogenic versions of Ras promote a CD24+to CD242transition and EMT in MCF10A and HMLE cell lines. (A) FACS analysis
of CD24 and CD44 markers in MCF10A cells infected either with a K-RasV12retroviral construct or the corresponding empty vector as a control. (B) Cell
morphology of MCF10A cells infected either with a K-RasV12retroviral construct or the corresponding empty vector as a control (upper panel) and
HMLE infected either with a H-RasV12retroviral construct or the corresponding empty vector as a control (lower panel). Images shown at 40x
magnification. (C) CD24+and CD242display epithelial and mesenchymal features, respectively. Top: Cell morphology. Bottom: Expression analysis as
assessed by western-blotting of epithelial (E-cadherin, b-catenin) and mesenchymal (vimentin, fibronectin) markers in two independent clones for
both cell populations. Images shown at 40x magnification.
Cancer Stem Cells and EMT
PLoS ONE | www.plosone.org4 August 2008 | Volume 3 | Issue 8 | e2888
Taken together, our observations demonstrate that, at least in
vitro, CD44+CD242/lowcells can originate from CD44lowCD24+
human mammary epithelial cells after aberrant activation of the
Ras/MAPK pathway. They also strongly suggest that the number
of CD242cells within a primary tumor reflects the sensitivity of
the cancer cells to EMT-inducing signals. Are these cells potential
‘‘tumor initiating cells’’? Although not tested in similar experi-
mental conditions, the tumorigenicity of CD242cells generated in
our experimental model appears to be significantly weaker than
the one of cancer stem cells originally isolated from human breast
cancer-derived malignant pleural effusions by the group of Clarke
. On this basis, two (non-exclusive) hypotheses can be proposed.
First, in the course of tumor progression, cancer cells with stem-like
capabilities can be generated from differentiated pre-malignant cells
by acquiring specific genetic alterations. As these somatic abnor-
malities also provide a growth advantage, the potential number of
such cells within a tumor might compensate for their limited stem-
like characteristics. Accordingly, these cells could act as an auxiliary
power source for tumor progression and metastasis. A second
hypothesis is that similar alterations couldinitially affectnormal stem
cells, giving rise to genuine cancer stem cells. The Weinberg group
recently reported that the tumorigenicity of experimentally trans-
formed mammary epithelial cells is highly dependent upon the cell
type of origin . When exposed to microenvironmental signals,
lending support to the notion of ‘‘mobile cancer stem cells’’ initially
proposed by T. Brabletz .
Of note, during the review process of our manuscript, the
connection between EMT and stem-like properties has also been
strongly supported by the Weinberg laboratory . Using different
EMT-inducers, they showed that the induction of EMT in
immortalized human mammary epithelial cells is associated with
the acquisition of stem-like characteristics. Additionally it was shown
that normal, as well as neoplastic breast stem-like cells, express
mesenchymal markers. These data further support our findings.
Materials and Methods
Cell culture, proliferation and mammosphere-formation
Human mammary epithelial cells were provided and cultured as
recommended by Lonza. HMEC-derivatives (kindly provided by
RA Weinberg) were cultured in 1:1 Dulbecco’s Modified Eagle’s
Medium (DMEM)/HAMF12 medium (Invitrogen) complemented
with 10% FBS (Cambrex), 100 U/ml penicillin-streptomycin
Figure 4. TGFb1 concomitantly promotes EMT and the CD24+to
CD242transition in CD24+HMLER cells. (A) Treatment of CD24+
HMLER with TGFb1 induces EMT, as assessed by morphology changes as
well as loss of E-cadherin (epithelial marker) and induction of vimentin
(mesenchymal marker), as assessed by immunofluorescence staining.
Images shown at 40x magnification. (B) Treatment of CD24+HMLER with
TGFb1 induces CD24+to CD242transition. FACS analysis of CD24 and
CD44 markers in CD24+cells, untreated or treated with TGFb1.
Figure 5. Oncogenic Ras and TGFb1 cooperate to promote the CD24+to CD242cells. HMLE cells were infected with an H-RasV12retroviral
expression construct or the empty vector (pBabe) as a control. Two days post-infection, experimental cells were treated with TGFb1. Percentage of
CD242cells was assessed at different times following infection. Error bars indicate +/2 standard deviation of triplicates.
Cancer Stem Cells and EMT
PLoS ONE | www.plosone.org5August 2008 | Volume 3 | Issue 8 | e2888
(Invitrogen), 2 mM L glutamine (Invitrogen), 10 ng/ml human
epidermal growth factor (EGF) (PromoCell), 0.5 mg/ml hydrocor-
tisone (Sigma) and 10 mg/ml insulin (Sigma) and treated with
10ng/ml recombinant TGFb1 (Peprotech) for 15 days.
was generously provided by the Ben-Ho Park’s laboratory and
cell line was maintained as recommended by the ATCC.
For mammosphere formation, after filtration through a 30 mm
pore filter, single-cells were plated at 105cells/ml in Corning 3261
ultra-low attachment culture dishes in the growth medium
described above. Primary cell spheres were enzymatically
dissociated with 0.05% trypsin for 15 minutes at 37uC to obtain
26106Phoenix cells were transfected by calcium-phosphate
precipitation with 10 mg of retroviral vector pBabe-H-RasV12.
48 hours post-transfection, the supernatant was collected, filtered,
supplemented with 4 mg/ml of polybrene (Sigma) and combined
with 106HMLE cells for 3 hours. Infected cells were selected after
48 hours with puromycin (0.5 mg/ml).
To measure anchorage-independent growth, cells were de-
tached with trypsin and resuspended in growth medium. Plates
were prepared with a coating of 0.75% agarose (Cambrex) in
growth medium and then overlaid with a suspension of cells in
0.45% agarose (56103cells/well). Plates were incubated for 3
weeks at 37uC and colonies were counted under microscope.
Identification and sorting of CD24+and CD44+cells were
performed using monoclonal anti-CD24-PE ML5 and anti-CD44-
FITC G44-26 antibodies (PharMingen), a FACScan (Becton
Dickinson) and a DIVA instrument (Becton Dickinson).
Animal maintenance and experiments were performed in
accordance with the animal care guidelines of the European Union
and French laws. Six-week old female Athymic Swiss nude mice (C.
River laboratories) were injected with 106CD242or CD24+cells
into a fat pad of mammary gland. Tumor growth was monitored
twice a week with callipers at the site of injection. Animals were
sacrificed as soon as tumor size reached 1.5 cm in diameter.
Cells were washed twice with phosphate buffered saline (PBS)
containing CaCl2 and then lysed in RIPA buffer. Protein
expression was examined by western blot using monoclonal anti-
E-cadherin clone 36 (Becton Dickinson), anti-b-catenin clone 14
(Becton Dickinson), anti-fibronectin FN-15 (Sigma), anti-vimentin
V9 (Dako), anti-b-actin AC-15 (Sigma) antibodies for primary
detection. Horseradish peroxidase-conjugated rabbit anti-mouse
antibody (Amersham) was used as a secondary antibody. Western-
blots were revealed using an ECL detection kit (Amersham).
About 104cells were seeded on 4-well Lab-TekII chamber slide.
After TGFb1 treatment, the cells were washed with PBS twice,
fixed in 3% parformaldehyde (Sigma) and permeabilized in 0.1%
Triton 100X (Sigma) in PBS buffer at 4uC for 30 minutes. The
cells were then washed 3 times with PBS and incubated with
blocking solution (10% horse serum in PBS). The cells were then
incubated with the primary antibodies anti-E-cadherin clone 36
(Becton Dickinson), or anti-vimentin V9 (Dako) overnight at 4uC.
The cells were washed 3 times in PBS and incubated with the
appropriate secondary antibodies (Dako) for 1 hour at room
temperature. Finally the cells were washed 3 times in PBS and
incubated with Hoechst (Sigma) for 5 minutes. The slides were
washed extensively with PBS and mounted with Fluoromount-G
(SouthernBiotech). All matched samples were photographed
(control and test) using immunofluorescence microscope and
identical exposure times.
The authors thank Marie-Dominique Reynaud and PROFESSional
Editors & Writers for help in the manuscript preparation. We extend
special thanks to Isabelle Treilleux (Department of Anatomo-Pathology,
Centre Le ´on Be ´rard, Lyon), Fre ´de ´rique Fauvet for technical assistance and
Isabelle Durand for help in FACS analysis. We would like to thank Dr
Robert A. Weinberg and the Dr Ben Ho Park for providing HMEC-
derivatives and Ras-transformed MCF10A cell lines, respectively.
Conceived and designed the experiments: APM SA AP. Performed the
experiments: APM ML CT. Analyzed the data: APM AP. Contributed
reagents/materials/analysis tools: ML CT. Wrote the paper: APM GH SA
1. Pardal R, Clarke MF, Morrison SJ (2003) Applying the principles of stem-cell
biology to cancer. Nat Rev Cancer 3: 895–902.
2. Vermeulen L, Sprick MR, Kemper K, Stassi G, Medema JP (2008) Cancer stem
cells-old concepts, new insights. Cell Death Differ.
3. Ponti D, Costa A, Zaffaroni N, Pratesi G, Petrangolini G, et al. (2005) Isolation
and in vitro propagation of tumorigenic breast cancer cells with stem/progenitor
cell properties. Cancer Res 65: 5506–5511.
4. Bonnet D, Dick JE (1997) Human acute myeloid leukemia is organized as a
hierarchy that originates from a primitive hematopoietic cell. Nat Med 3:
5. Miyamoto T, Weissman IL, Akashi K (2000) AML1/ETO-expressing
nonleukemic stem cells in acute myelogenous leukemia with 8;21 chromosomal
translocation. Proc Natl Acad Sci U S A 97: 7521–7526.
6. Cobaleda C, Gutierrez-Cianca N, Perez-Losada J, Flores T, et al. (2000) A
primitive hematopoietic cell is the target for the leukemic transformation in
human philadelphia-positive acute lymphoblastic leukemia. Blood 95:
7. Huntly BJ, Shigematsu H, Deguchi K, Lee BH, Mizuno S, et al. (2004) MOZ-
TIF2, but not BCR-ABL, confers properties of leukemic stem cells to committed
murine hematopoietic progenitors. Cancer Cell 6: 587–596.
8. Cozzio A, Passegue E, Ayton PM, Karsunky H, Cleary ML, et al. (2003) Similar
MLL-associated leukemias arising from self-renewing stem cells and short-lived
myeloid progenitors. Genes Dev 17: 3029–3035.
9. So CW, Karsunky H, Passegue E, Cozzio A, Weissman IL, et al. (2003) MLL-
GAS7 transforms multipotent hematopoietic progenitors and induces mixed
lineage leukemias in mice. Cancer Cell 3: 161–171.
10. Krivtsov AV, Twomey D, Feng Z, Stubbs MC, Wang Y, et al. (2006)
Transformation from committed progenitor to leukaemia stem cell initiated by
MLL-AF9. Nature 442: 818–822.
11. Lapidot T, Sirard C, Vormoor J, Murdoch B, Hoang T, et al. (1994) A cell
initiating human acute myeloid leukaemia after transplantation into SCID mice.
Nature 367: 645–648.
12. Turhan AG, Lemoine FM, Debert C, Bonnet ML, Baillou C, et al. (1995) Highly
purified primitive hematopoietic stem cells are PML-RARA negative and generate
nonclonal progenitors in acute promyelocytic leukemia. Blood 85: 2154–2161.
13. Holyoake TL, Jiang X, Drummond MW, Eaves AC, Eaves CJ (2002)
Elucidating critical mechanisms of deregulated stem cell turnover in the chronic
phase of chronic myeloid leukemia. Leukemia 16: 549–558.
14. Nilsson L, Strand-Grundstrom I, Arvidsson I, Jacobsson B, Hellstrom-
Lindberg E, et al. (2000) Isolation and characterization of hematopoietic
Cancer Stem Cells and EMT
PLoS ONE | www.plosone.org6August 2008 | Volume 3 | Issue 8 | e2888
progenitor/stem cells in 5q-deleted myelodysplastic syndromes: evidence for
involvement at the hematopoietic stem cell level. Blood 96: 2012–2021.
15. Al-Hajj M, Wicha MS, Ito-Hernandez A, Morrison SJ, Clarke MF (2003)
Prospective identification of tumorigenic breast cancer cells. Proc Natl Acad
Sci U S A 100: 3983–3988.
16. Singh SK, Hawkins C, Clarke ID, Squire JA, Bayani J, et al. (2004)
Identification of human brain tumour initiating cells. Nature 432: 396–401.
17. O’Brien CA, Pollett A, Gallinger S, Dick JE (2007) A human colon cancer cell
capable of initiating tumour growth in immunodeficient mice. Nature 445:
18. Collins AT, Berry PA, Hyde C, Stower MJ, Maitland NJ (2005) Prospective
identification of tumorigenic prostate cancer stem cells. Cancer Res 65:
19. Suetsugu A, Nagaki M, Aoki H, Motohashi T, Kunisada T, et al. (2006)
Characterization of CD133+ hepatocellular carcinoma cells as cancer stem/
progenitor cells. Biochem Biophys Res Commun 351: 820–824.
20. Yin S, Li J, Hu C, Chen X, Yao M, et al. (2007) CD133 positive hepatocellular
carcinoma cells possess high capacity for tumorigenicity. Int J Cancer 120:
21. Eramo A, Lotti F, Sette G, Pilozzi E, Biffoni M, et al. (2008) Identification and
expansion of the tumorigenic lung cancer stem cell population. Cell Death Differ
22. Uchida N, Buck DW, He D, Reitsma MJ, Masek M, et al. (2000) Direct isolation
of human central nervous system stem cells. Proc Natl Acad Sci U S A 97:
23. Dalerba P, Dylla SJ, Park IK, Liu R, Wang X, et al. (2007) Phenotypic
characterization of human colorectal cancer stem cells. Proc Natl Acad Sci U S A
24. Ricci-Vitiani L, Lombardi DG, Pilozzi E, Biffoni M, Todaro M, et al. (2007)
Identification and expansion of human colon-cancer-initiating cells. Nature 445:
25. Al-Hajj M, Clarke MF (2004) Self-renewal and solid tumor stem cells. Oncogene
26. Dontu G, Abdallah WM, Foley JM, Jackson KW, Clarke MF, et al. (2003) In
vitro propagation and transcriptional profiling of human mammary stem/
progenitor cells. Genes Dev 17: 1253–1270.
27. Elenbaas B, Spirio L, Koerner F, Fleming MD, Zimonjic DB, et al. (2001)
Human breast cancer cells generated by oncogenic transformation of primary
mammary epithelial cells. Genes Dev 15: 50–65.
28. Liu R, Wang X, Chen GY, Dalerba P, Gurney A, et al. (2007) The prognostic
role of a gene signature from tumorigenic breast-cancer cells. N Engl J Med 356:
29. Konishi H, Karakas B, Abukhdeir AM, Lauring J, Gustin JP, et al. (2007)
Knock-in of mutant K-ras in nontumorigenic human epithelial cells as a new
model for studying K-ras mediated transformation. Cancer Res 67: 8460–8467.
30. Shipitsin M, Campbell LL, Argani P, Weremowicz S, Bloushtain-Qimron N, et
al. (2007) Molecular definition of breast tumor heterogeneity. Cancer Cell 11:
31. Thiery JP (2002) Epithelial-mesenchymal transitions in tumour progression. Nat
Rev Cancer 2: 442–454.
32. Hay JG, McElvaney NG, Herena J, Crystal RG (1995) Modification of nasal
epithelial potential differences of individuals with cystic fibrosis consequent to
local administration of a normal CFTR cDNA adenovirus gene transfer vector.
Hum Gene Ther 6: 1487–1496.
33. Christiansen JJ, Rajasekaran AK (2006) Reassessing epithelial to mesenchymal
transition as a prerequisite for carcinoma invasion and metastasis. Cancer Res
34. Oft M, Peli J, Rudaz C, Schwarz H, Beug H, et al. (1996) TGF-beta1 and Ha-
Ras collaborate in modulating the phenotypic plasticity and invasiveness of
epithelial tumor cells. Genes Dev 10: 2462–2477.
35. Gotzmann J, Mikula M, Eger A, Schulte-Hermann R, Foisner R, et al. (2004)
Molecular aspects of epithelial cell plasticity: implications for local tumor
invasion and metastasis. Mutat Res 566: 9–20.
36. Massague J (2007) Sorting out breast-cancer gene signatures. N Engl J Med 356:
37. Sheridan C, Kishimoto H, Fuchs RK, Mehrotra S, Bhat-Nakshatri P, et al.
(2006) CD44+/. Breast Cancer Res 8: R59.
38. Ince TA, Richardson AL, Bell GW, Saitoh M, Godar S, et al. (2007)
Transformation of different human breast epithelial cell types leads to distinct
tumor phenotypes. Cancer Cell 12: 160–170.
39. Brabletz T, Jung A, Spaderna S, Hlubek F, Kirchner T (2005) Opinion:
migrating cancer stem cells-an integrated concept of malignant tumour
progression. Nat Rev Cancer 5: 744–749.
40. Mani SA, Guo W, Liao MJ, Eaton EN, Ayyanan A, et al. (2008) The epithelial-
mesenchymal transition generates cells with properties of stem cells. Cell 133:
Cancer Stem Cells and EMT
PLoS ONE | www.plosone.org7 August 2008 | Volume 3 | Issue 8 | e2888