Alonso-Alconada et al.

Molecular Cancer 2014, 13:223

http://www.molecular-cancer.com/content/13/1/223

RESEARCH Open Access

Molecular profiling of circulating tumor cells links

plasticity to the metastatic process in endometrial
cancer
Lorena Alonso-Alconada1, Laura Muinelo-Romay1, Kadri Madissoo2,3, Antonio Diaz-Lopez4, Camilla Krakstad2,3,
Jone Trovik2,3, Elisabeth Wik3,5, Dharani Hapangama6, Lieve Coenegrachts7, Amparo Cano4, Antonio Gil-Moreno8,
Luis Chiva9, Juan Cueva1, Maria Vieito1, Eugenia Ortega10, Javier Mariscal1, Eva Colas8, Josep Castellvi8,
Maite Cusido11, Xavier Dolcet10, Hans W Nijman12, Tjalling Bosse13, John A Green6, Andrea Romano14,
Jaume Reventos8,15, Rafael Lopez-Lopez1, Helga B Salvesen2,3, Frederic Amant7, Xavier Matias-Guiu10,
Gema Moreno-Bueno4,9, and Miguel Abal1* on behalf of ENITEC Consortium

Abstract
Background: About 20% of patients diagnosed with endometrial cancer (EC) are considered high-risk with
unfavorable prognosis. In the framework of the European Network for Individualized Treatment in EC (ENITEC),
we investigated the presence and phenotypic features of Circulating Tumor Cells (CTC) in high-risk EC patients.
Methods: CTC isolation was carried out in peripheral blood samples from 34 patients, ranging from Grade 3 Stage
IB to Stage IV carcinomas and recurrences, and 27 healthy controls using two methodologies. Samples were
subjected to EpCAM-based immunoisolation using the CELLection™ Epithelial Enrich kit (Invitrogen, Dynal) followed
by RTqPCR analysis. The phenotypic determinants of endometrial CTC in terms of pathogenesis, hormone receptor
pathways, stem cell markers and epithelial to mesenchymal transition (EMT) drivers were asked. Kruskal-Wallis
analysis followed by Dunn’s post-test was used for comparisons between groups. Statistical significance was set
at p < 0.05.
Results: EpCAM-based immunoisolation positively detected CTC in high-risk endometrial cancer patients. CTC
characterization indicated a remarkable plasticity phenotype defined by the expression of the EMT markers ETV5,
NOTCH1, SNAI1, TGFB1, ZEB1 and ZEB2. In addition, the expression of ALDH and CD44 pointed to an association
with stemness, while the expression of CTNNB1, STS, GDF15, RELA, RUNX1, BRAF and PIK3CA suggested potential
therapeutic targets. We further recapitulated the EMT phenotype found in endometrial CTC through the
up-regulation of ETV5 in an EC cell line, and validated in an animal model of systemic dissemination the propensity
of these CTC in the accomplishment of metastasis.
Conclusions: Our results associate the presence of CTC with high-risk EC. Gene-expression profiling characterized a
CTC-plasticity phenotype with stemness and EMT features. We finally recapitulated this CTC-phenotype by
over-expressing ETV5 in the EC cell line Hec1A and demonstrated an advantage in the promotion of metastasis in
an in vivo mouse model of CTC dissemination and homing.
Keywords: High-risk endometrial carcinomas, Circulating tumor cells, Epithelial to mesenchymal transition,
Stem cell, ETV5

* Correspondence: [email protected]
1
Translational Medical Oncology; Health Research Institute of Santiago (IDIS),
SERGAS, Trav. Choupana s/n 15706, Santiago de Compostela, Spain
Full list of author information is available at the end of the article

© 2014 Alonso-Alconada et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the
Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use,
distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public
Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this
article, unless otherwise stated.
Alonso-Alconada et al. Molecular Cancer 2014, 13:223 Page 2 of 10
http://www.molecular-cancer.com/content/13/1/223

Introduction Results
Endometrial carcinomas, the most common tumors of Assessment of CTC in the blood samples from high-risk
the female genital tract, are usually diagnosed at an early EC patients
stage with uterine-confined disease and an overall fa- Immunoisolation of CTC from peripheral blood samples
vorable prognosis. However, up to 20% of endometrial has been performed with magnetic beads coated with
carcinomas present as aggressive neoplasms such as EpCAM antibodies. We thus first confirmed the posi-
high-grade or deeply invasive lesions, at substantial risk tivity for EpCAM expression in the corresponding pri-
of recurrence and death [1]. Propagation of aggressive mary carcinomas of a representative sample of patients
tumor cells from the primary lesion is a key event in the included in the study (Figure 1A). Secondly, we investi-
process of metastasis and a challenge in oncology. In gated the presence and quantified the amounts of CTC
endometrial cancer, myometrial infiltration, lymph node in a series of 34 EC patients ranging from Grade 3 Stage
involvement, and lymphovascular space invasion are IB carcinomas to metastatic Stage IV carcinomas and re-
current clinical parameters defining the probability of currences (see global clinical descriptions in Table 1).
recurrent disease. Nevertheless, early dissemination of CTC were immunoisolated with EpCAM-dynabeads
tumor cells is usually undetectable in patients by con- from EDTA-BD Vacutainer 7.5 ml blood collection tube.
ventional histopathological examination or by standard Upon RNA extraction and pre-amplification, we eva-
imaging techniques. Recently, immunocytochemical and luated the expression levels of GAPDH as a marker of
molecular assays have been developed for the specific cellularity, which includes both CTC and unspecific
detection of metastatic tumor cells at a cellular level in blood cells, normalized to the background of CD45 ex-
lymph nodes, peripheral blood or bone marrow, prior to pression as specific marker for cells of hematopoietic
the manifestation of metastasis. Tumor-cell dissemi- origin [7]. As shown, GAPDH levels were significantly
nation can proceed at an early stage of tumor develop- higher in the group of patients compared to controls
ment [2], and detecting circulating tumor cells (CTC) (Figure 1B; with 30 high-risk EC patients below control
has clinical value in the monitoring and the outcome of upper threshold and 3 controls above lower patients’
metastatic disease. CTC analysis represents an attractive threshold), while CD45 did not present differences bet-
candidate for liquid biopsy in cancer [3]. Clinically, the ween both groups (Figure 1C), indicating (i) the pre-
presence of CTC above a threshold may have a signifi- sence of an extra population of cells isolated from the
cant adverse impact on survival. Likewise, changes on blood of high-risk EC patients and (ii) the unspecific
CTC quantification during treatment can reflect prog- background resulting from the process of immunoisola-
nostic significance, the future challenge being whether tion was similar in the group of patients and controls.
treatment decision-making should be impacted by CTC The presence of CTC in high-risk endometrial cancer
levels [4]. patients was further confirmed with the technology that
The increasing interest in CTC at the clinical setting is received to date Food and Drug Administration (FDA)
resulting in the development of a number of innovative clearance for the monitoring of metastatic breast, colorec-
technologies that include immunoenrichment, micro- tal, and prostate cancer, the CellSearch System (Janssen
fluidics and filtration devices, combined with semiauto- Diagnostics, SouthRaritan, NJ, USA), which combines im-
mated microscopy or PCR-based detection systems [5]. munoenrichment and immunofluorescence for the de-
We have recently demonstrated that the combination of tection of CTC [8-10] (Additional file 1). Globally, these
CTC EpCAM-based immunoisolation, followed by ac- results demonstrated in parallel the presence of CTC in
curate extraction and pre-amplification of RNA from high-risk EC patients.
very small number of CTC, provided with a highly sen-
sitive approach to profiling the metastatic tumor cell Gene-expression analysis in immunoisolated CTC from EC
population in a group of colorectal cancer patients [6]. patients highlights a plasticity phenotype
In the present study, we adopted a similar approach in Once we confirmed their presence, we explored the gene-
high-risk EC patients. CTC immunoisolation plus pro- expression profile of CTC in the samples from high-risk
filing of a number of genes related to key events in the EC patients upon EpCAM-based immunoisolation and
process of metastasis in EC provided us with an over- RNA extraction and pre-amplification. For this, we
view of the biology of endometrial CTC. In addition to analyzed genes of signaling pathways previously reported
analyze in immunoisolated CTC the expression of a to be related to EC (BRAF, CTNNB1, ERBB2, FGFR2,
number of genes involved in signaling pathways related GDF15, IDO, MTOR, P53, PIK3CA, PTEN, PTGS2,
to EC, hormone pathways, stem cell features and epi- RUNX1, RELA, STMN1, TERT, VIL1, ZWINT), hormone
thelial to mesenchymal transition (EMT) markers, we pathways (CYP19, ESR1, ESR2, GPER, HSD17B1, PGR,
evaluated the efficiency of CTC quantification and its STS, TFF1), stem cell features (ALDH, CD133, CD44), and
correlation with clinical parameters. EMT related markers (ETV5, LOXL2, NOTCH1, SNAI1,
Alonso-Alconada et al. Molecular Cancer 2014, 13:223 Page 3 of 10
http://www.molecular-cancer.com/content/13/1/223

Figure 1 Immunoisolation of CTC in high-risk EC patients. (A) Representative primary carcinoma of a patient included in the study, demonstrating
positivity for membrane EpCAM staining in epithelial tumor cells. (B) GAPDH expression levels normalized to CD45 in CTC isolated from the group
of controls (white box, n = 27), as background of unspecific immunoisolation, and from the group of high-risk EC patients (grey box, n = 34)
(Mann–Whitney test, ***p < 0.001). (C) CD45 expression levels in CTC isolated from controls and patients; similar expression denoted equivalent
degree of unspecific immunoisolation.

Table 1 Clinical and pathologic characteristics of high-risk EC patients included in the gene-expression profiling of CTC
Feature n (%) Feature n (%)
Age (years)
Mean 69
FIGO Stage Lymphovascular invasion
I 15 (44.15) Positive 6 (17.6)
II 2 (5.8) Negative 15 (44.1)
III 11 (32.35) Unknown 13 (38.2)
IV 6 (17.64) Lymph node metastasis
Histology Positive 12 (35.3)
Endometrioid 19 (55.9) Negative 20 (58.8)
Serous 10 (29.4) Unknown 2 (5.9)
Clear cell 5 (14.7) Recurrence
Grade No 18 (52.9)
Well differentiated (G1) 4 (11.8) Yes 15 (44.1)
Moderately differentiated (G2) 6 (17.6) Unknown 1 (2.9)
Poorly differentiated (G3) 21 (61.8) First treatment
Unknown 3 (8.8) Radiotherapy 4 (11.8)
Myometrial invasion Chemotherapy 7 (20.6)
<50% 13 (38.2) None 17 (50)
>50% 18 (52.9) Unknown 6 (17.6)
Unknown 3 (8.8)
Unknown values were considered as missing and therefore excluded from the statistical analysis.
Alonso-Alconada et al. Molecular Cancer 2014, 13:223 Page 4 of 10
http://www.molecular-cancer.com/content/13/1/223

TGFB1, ZEB1, ZEB2) (Additional file 2). Initially, where in Among genes related to EC pathogenesis, we found sta-
a preliminary set of 6 patients and 6 controls no amplified tistical increased expression in the group of Stage III-IV
signal could be observed either in patients or in controls carcinomas and recurrences compared with healthy
probably due to non-detectable levels of expression, genes controls for BRAF, for the Wnt pathway component
were excluded from further evaluation. This primary CTNNB1, for the TGF-beta superfamily cytokine GDF15,
screen resulted in exclusion of CD133, GPER, HSD17B1, for PIK3CA, for NF-κB family member RELA, and for the
PGR, and TERT (data not shown). RUNX transcription factor family member RUNX1,
We next analyzed the expression levels of the re- (Figure 2A). Regarding genes from hormone pathways, we
maining genes in the whole set of patients and controls, found a significant increased expression of STS in ad-
and identified those genes with a significant expression vanced stage cases compared with controls (Figure 2B).
in CTC from the group of patients compared to the Increased expression in both ALDH and CD44 as genes
background of unspecific isolation from the controls. related to stem cell features was observed in CTC immu-
These genes are considered to characterize the popula- noisolated from Stages III and IV high-risk EC patients
tion of CTC in EC. We further grouped those high-risk and recurrences compared with controls (Figure 2C). The
patients presenting FIGO Stages I-II and those with correlation between the levels of expression of these can-
FIGO Stages III-IV and recurrences, as stages presenting didate biomarkers and the clinicopathological features of
no or local cell tumor spread and those presenting patients included in the study, indicated that the expres-
systemic dissemination, respectively. Overall, major sig- sion of BRAF, PIK3CA, RELA, RUNX1 (all involved in
nificances were found between the group of healthy con- endometrial carcinogenesis) and CD44 (stem cell marker)
trols and FIGO Stages III-IV and recurrences, consistent was increased in CTC isolated from patients with the
with a gradual increased presence of CTC as the disease tumor infiltrating more than 50% of the myometrium
disseminates; modest increases in a reduced number of (n = 18) compared with those isolated from patients
genes in FIGO Stages I-II that might suggest the pre- having myometrial invasion below 50% of the myome-
sence of CTC in early stages has to be interpreted trium (n = 13) (Additional file 3).
with caution due to the limited number of patients More interestingly, almost the complete set of genes
(Figure 2A). associated with EMT (ETV5, NOTCH1, SNAI1, TGFB1,

Figure 2 Gene expression profiling in endometrial CTC. (A) Significant expression levels of genes involved in signaling pathways reported
altered in EC, (B) hormone pathways and (C) stem cell features, in CTC from high-risk EC patients compared to the background of unspecific
immunoisolation. White boxes represent the gene expression levels in the group of healthy controls, light grey boxes those corresponding to
FIGO Stages I and II EC patients while dark grey boxes those corresponding to FIGO Stages III-IV EC patients and recurrences. (Kruskal-Wallis test,
*p < 0.05; **p < 0.01; ***p < 0.001).
Alonso-Alconada et al. Molecular Cancer 2014, 13:223 Page 5 of 10
http://www.molecular-cancer.com/content/13/1/223

ZEB1 and ZEB2) were found to be specifically expressed Figure 4A, an increase in ETV5 expression resulted in
in CTC from EC patients when compared to unspecific the concomitant up-regulation of genes related to EMT,
background from controls (Figure 3), with LOXL2 as observed in CTC. These results suggested that at least
achieving a value close to significance (p = 0.08; data not part of the plasticity phenotype observed in CTC in EC
shown). We also found statistical increased expression patients might be associated with the up-regulation of
levels in the EMT marker ZEB2, together with RUNX1, ETV5 during myometrial invasion and confirmed the
when we compared high-risk EC patients presenting re- role of ETV5 as a master regulator of EMT.
currences (n = 15) or not (n = 18) (Additional file 3). We further investigated the impact of ETV5 up-
Likewise, when we assessed the panel of CTC signifi- regulation and the consequent EMT phenotype in a
cantly expressed genes in a small series of matched pri- mice model that reproduced the systemic dissemination,
mary endometrial carcinomas and affected lymph nodes homing and generation of micrometastasis associated
(n = 6), the expression of the EMT related gene ZEB2 with CTC. For this, we compared the extent of metasta-
demonstrated the best performance among a global ten- sis resulting from the intracardiac injection of both naïve
dency for an increased expression in lymph node metas- Hec1A cells (Figure 4B, upper panels), and ETV5 over-
tasis compared to primary lesions, (p = 0.09; Additional expressing Hec1A cells (Figure 4B, lower panels), further
file 4). No significant differences were found for any spe- modified to constitutively express the luciferase reporter
cific histology subtype regarding EMT markers, but we gene and allow vital bioluminescent imaging of dis-
cannot exclude whether this could be due to the limited seminating tumor cells. As shown in Figure 4B, the up-
number of samples in the study or to the absence of dif- regulation of ETV5 and the consequent recapitulation of
ferences in disseminated disease. We are currently in- the CTC-plasticity phenotype resulted in a more dra-
creasing the number of paired matched samples in order matic cell dissemination and metastatic spread, both in
to confirm these data. Overall, gene profiling on CTC terms of pattern of metastasis localization as well as the
isolated from high-risk EC patients suggested a plasticity number of metastasis foci (Figure 4C, D). In addition to
phenotype characterized by genes of stem cell features reinforce the role of ETV5 during the initial events of
and EMT, with potential in the clinical management of metastasis in endometrial cancer, these results link the
high-risk EC patients. promotion of a plasticity phenotype in CTC with their
capacity to metastasize.
ETV5 recapitulates the EMT phenotype characterizing CTC
and impacts metastasis in an EC mouse model Discussion
ETV5 transcription factor, one of the candidate genes We present in this study evidences for the presence of
characterizing the CTC-plasticity phenotype, plays a CTC in high-risk EC patients, and further characterized
major role in EC as a regulator of EMT [11,12]. This led a molecular CTC-phenotype associated with plasticity
us to further asses the role of ETV5 in endometrial CTC and stemness features. The major clinical relevance of
and evaluate whether the invasive phenotype promoted CTC is that the early detection in patients could be of
by ETV5 during myometrial infiltration recapitulated use for the identification of candidate subjects needing
the CTC phenotype observed in patients. To this end, additional systemic therapies after the resection of the
the naïve EC cell line Hec1A and the cell line Hec1A primary tumor. Although the aim of these therapies is
over-expressing ETV5 were used. As can be observed in the prevention of metastasis, the selection of patients is

Figure 3 Plasticity phenotype characterizes CTC in EC. Almost all genes related to EMT assessed in CTC from high-risk EC patients presented
significant expression compared to the background of unspecific immunoisolation. White boxes represent the gene expression levels in the group
of healthy controls, light grey boxes those corresponding to FIGO Stages I and II EC patients while dark grey boxes those corresponding to FIGO
Stages III-IV EC patients and recurrences. (Kruskal-Wallis test, *p < 0.05; **p < 0.01; ***p < 0.001).
Alonso-Alconada et al. Molecular Cancer 2014, 13:223 Page 6 of 10
http://www.molecular-cancer.com/content/13/1/223

effects associated with un-necessary treatments [13]. In

addition, the sequential assessment of CTC levels during
treatment could provide information at early stages
about the therapeutic efficacy of drugs. Likewise, the
elimination of these CTC could represent an inter-
mediate endpoint in clinical trials with antitumor drugs.
Furthermore, the molecular and functional characte-
rization of CTC will be determinant in the discovery of
new molecular tumor markers and in the development
of therapies specific for the process of tumor dissemi-
nation and metastasis. Overall, CTC represent a potent
and promising tool in oncology and, although the cli-
nical role of CTC as a prognosis factor is being recog-
nized, there is still a need for more advanced and precise
techniques of detection and robust clinic-pathologic cor-
relations [14].
From our results, we can conclude that EpCAM-based
immune-enrichment followed by RT-qPCR analysis is a
reliable method to effectively isolate CTC from high-risk
EC patients and to potentially distinguish high-risk from
low-risk patients. Previous efforts to analyze and evaluate
CTC in EC included the assessment of a six gene panel in
blood samples [15]. Although this methodology represents
a semi-quantitative evaluation of CTC, it demonstrated an
added value compared to the gold standard CellSearch
technology for the positive detection of CTC (see
Additional file 1). In addition to the clinical utility of CTC
as a surrogate marker in the management of high-risk and
metastatic cancer patients, the challenge stands on the
possibility of a therapeutic approach targeting these meta-
static CTC with the aim of controlling and/or eradicating
the source of recurrences. The advantage of combining
CTC immunoisolation and RT-qPCR analysis consists in
the possibility to identify and characterize biomarkers spe-
cific of this subpopulation of metastatic cells. This ap-
proach has the limitation to determine whether the
Figure 4 ETV5 recapitulates the EMT phenotype found in CTC increased levels in biomarker expression upon CTC
and the metastasis potential in an EC mouse model. (A) CTC-gene immunoisolation is related to an enhanced expression or
expression profiling in Hec1A and Hec1A-ETV5 cell lines by RT-qPCR. to an augmentation in the number of CTC. Nevertheless,
The results were represented as the fold change in gene expression
relative to GAPDH gene expression (2-ΔΔCt). (B) Representative
in addition to their potential in terms of diagnosis/progno-
luminiscence examples of athymic nude mice inoculated with Hec1A sis and follow-up of patients, these biomarkers provide
(upper panels) or Hec1A-ETV5 (lower panels) cells by intracardiac with phenotypic clues on the biology of CTC that may be
injection. Luminescence images were acquired for 1 min at ventral determinant in the identification of new therapeutic strat-
(left image) and dorsal (right image) positions. (C) Extent of egies aiming to specifically control and/or eradicate the
dissemination evaluated as number of metastasis and (D) as
luminescence quantification of metastasis.
metastatic dissemination in EC [16]. From a clinical per-
spective, the management of metastatic EC has been
recently considered as a poly-chemotherapy adjuvant regi-
nowadays based on the statistical risk of recurrences, men for patients with high risk of relapse, and as a pal-
which is not accurate in terms of over-treatment of pa- liative regimen for patients with disseminated disease or
tients with toxic agents or therapeutic procedures caus- with extrapelvic recurrence not responding to hormone
ing serious side effects, in addition to economic costs. treatment. In addition, endometrial carcinomas are
An improved ability to target surgical and systemic the- considered as chemoresistant and the most active drugs
rapies to well selected high-risk patient populations will (platinum salts, doxorubicin, anthracyclines and pacli-
increase the likelihood of benefits and decrease the side taxel) present relative rate responses ranging from 25% in
Alonso-Alconada et al. Molecular Cancer 2014, 13:223 Page 7 of 10
http://www.molecular-cancer.com/content/13/1/223

monotherapy to 57% in poly-chemotherapy, with median support the hypothesis of a CTC phenotype with plasticity
survival of 12–15 months. In this scenario, chemotherapy and stemness features with the capacity to develop metas-
has shown limited utility and there is a clear need for the tasis [29]. This concept linking EMT and stem cell features
development of new rationale therapies focused on metas- in the process of tumor dissemination has also been ad-
tasis [17]. To this regard, our profiling of endometrial dressed in EC [30], in association with a micro-RNA sig-
CTC has pointed out at a number of pathways relevant to nature of EMT in endometrial carcinosarcoma mainly
the metastatic process and that could be targeted. For in- represented by the down-regulation of members of the
stance, the expression of STS, which is associated with miR-200 family [31]. Remarkably, the balanced expression
higher availability of estrogens in tumor cells [18], can of ZEB factors and miR-200 is considered as a molecular
provide a growth advantage to CTC and its inhibition can motor of cellular plasticity, in particular is a driving force
be used to block such event and decrease the risk of recur- for cancer progression towards metastasis by controlling
rences. A second important marker highly expressed in the state of cancer stem cells [32,33]. The results we ob-
CTC was PIK3CA, which confirms the role of this kinase tained both in CTC and in paired carcinoma and lymph-
pathway as a potential target in high-risk and metastatic node tissue samples demonstrating the potency of ZEB2
disease [19-21]. within the CTC-phenotype reinforce the need of future
A main feature observed in the molecular profiling of investigations examining stem-like features in CTC to ob-
CTC in EC corresponded to the EMT phenotype, with tain relevant information about this cancer subpopulation
almost all analyzed genes related to plasticity being sig- responsible of metastases.
nificantly expressed in CTC (ETV5, NOTCH1, SNAI1, Finally, it should be note that although EpCAM ex-
TGFB1, ZEB1, ZEB2). EMT is a dynamic process pression was found consistent in primary endometrial
whereby epithelial cells lose polarity and cell-cell con- carcinomas, its proposed modulation and eventual loss
tacts, undergo dramatic remodeling of the cytoskeleton, during EMT adds controversy to the efficiency of en-
acquire a migratory phenotype and a mesenchymal-like richment of CTC owning a plasticity phenotype [34]. To
gene expression program. Both invasion and metastasis this regard, we analyzed the expression of EpCAM both
may be critically dependent on the acquisition by the in- in the epithelial endometrial cancer cell line HEC1A
cipient cancer cell of EMT features [22,23]. Interestingly, and its mesenchymal counterpart Hec1A-ETV5, and
both EMT and the PIK3CA pathway have been closely found similar levels of EpCAM expression irrespective
linked in the promotion of metastasis, particularly in EC of their EMT phenotype (Additional file 5). Moreover,
[24,25]. Our results also reinforce a role for ETV5 in the EpCAM-based immunoisolation of these HEC1A and
process of EMT and EC dissemination [11]. ETV5 up- HEC1A-ETV5 cells lines rendered similar efficiencies
regulation in Hec1A cells recapitulated in vitro the (75% versus 64%, respectively). From these results, it
plasticity phenotype found in high-risk patients, and seems reasonable to speculate that cells detaching from
demonstrated an advantage in the promotion of metas- the primary lesion and incorporating into the blood
tasis in an in vivo mouse model that mimic CTC dis- stream recapitulate a metastable epithelial–mesenchy-
semination and homing. mal phenotype that may be maintained during their way
Likewise, and concerning ALDH and CD44 as stem-cell to those distant sites where this CTC will home and end
genes identified in this profiling, we observed a concor- up in the generation of micrometastasis. The dis-
dance between the presence of endometrial CTC and sociation of tumor cells from the epithelial layer and the
recurrent disease. Whether these CTC include a subpo- penetration through the basement membrane into the
pulation of tumor cells with a stem cell-like phenotype adjacent connective tissue, are the initial events in the
(Cancer Initiating Cells [26]) or whole CTC population multistep process that characterizes metastasis [35]. We
must be considered responsible for recurrences with more are additionally conducting further studies with other
or less efficiency, has yet to be addressed. In addition to immunoisolating antigens.
its plasticity phenotype, CTC triggering micrometastasis
in the target organs must own the capacity to survive in Conclusions
the blood flow, to home and to regenerate a tumor mass We present evidences for the presence of CTC in high-
with similar characteristics as the primary lesion in the tis- risk EC patients, and we propose a CTC-phenotype in
sue recipient of metastasis. Recent studies on CTC in EC associated with plasticity and stem cell features. Al-
breast cancer have demonstrated that a subset of isolated though this multicentre study conducted within the
CTC express stem cell markers such as those analysed in framework of ENITEC has the limitation of the number
our study [27,28]. The evidence that tumor dissemination of samples, these promising data offer the opportunity
to the blood circulation is an early event and that the to design new therapeutic strategies targeting metastatic
process of metastasis is an ineffective process with only a disease, and a larger prospective study is aimed for the
small number of CTC ending up in micrometastasis, validation of the CTC-phenotype in high-risk EC.
Alonso-Alconada et al. Molecular Cancer 2014, 13:223 Page 8 of 10
http://www.molecular-cancer.com/content/13/1/223

Methods subjected to pre-amplification with TaqMan®PreAmp

Patient samples Master Mix kit (Applied Biosystems, Foster City, CA, USA)
Peripheral blood samples collected just before initiation of for 14 reaction cycles before proceeding to RTqPCR,
treatment from 34 EC patients and 27 controls were proc- to provide with optimal detection rates. TaqMan Gene
essed for CTC immunoisolation and accurate RNA extrac- Expression Assays (Applied Biosystems, Foster City, CA,
tion as described [7]. Patients participating in the study USA) for 35 selected genes (Additional file 2; plus GAPDH
were surgically staged according to FIGO and recruited as housekeeping gene and CD45 as a marker of non-
between March 2012 - October 2013 in Vall d’Hebron Uni- specific isolation) were used to measure the gene ex-
versity Hospital (Barcelona, Spain), University Hospital of pression levels in CTC isolated from patients in comparison
Santiago de Compostela (Santiago de Compostela, Spain), to the background of hematogenous cells unspecifically
Arnau de Vilanova Hospital (Lleida, Spain), MD-Anderson immunoisolated from the group of healthy controls. Values
Cancer Center Madrid (Madrid, Spain), Fundacion Dexeus were analyzed using StepOne Software v.2.1 (Applied
(Barcelona, Spain) and Haukeland University Hospital Biosystems, Foster City, CA, USA), normalized to CD45
(Bergen, Norway), and included high-risk endometrial car- and represented as (40–ΔCt), whereby ΔCt = duplicate
cinomas ranging from Grade 3 Stage IB carcinomas to mean (CtTARGET – CtCD45).
metastatic Stage IV carcinomas and recurrences (Table 1).
Control group included a set of 27 healthy women with ab-
Cell lines and cell culture
sence of a previous cancer episode and with an age range
The human endometrial carcinoma cell lines Hec1A
similar to patients. Informed consent approved by the
and Hec1A stably expressing the ETV5 transcription
relevant ethical committee was signed by all patients.
factor (Hec1A-ETV5) were maintained in McCoy’s 5A
In addition, fresh-frozen tissue from primary tumor and
Medium (Gibco, Grand Island, NY, USA) supplemented
paired affected lymphatic nodes from 6 EC patients were
with 10% FBS and 1% penicillin-streptomycin at 37°C in
provided by Tumor Bank of the Vall d’Hebron University
5% CO2, Hec1A-ETV5 cells further selected with
Hospital Biobank (Barcelona, Spain) with appropriate
Geneticin (500 μg/ml; Gibco, Grand Island, NY, USA).
ethics approval.
These cells were previously generated and thoroughly
characterized [11,12].
EpCAM immunohistochemistry
To monitor non-invasively tumor grafts of Hec1A and
EpCAM expression was checked in whole paraffin-
Hec1A-ETV5 cells, these cells were infected with lenti-
embedded sections of primary endometrial carcinomas
viruses bearing pLenti CMV V5-LUC Blast (w567-1)
from patients prospectively subjected to evaluation of
(Addgene, Cambridge, MA, USA) to constitutively ex-
CTC. Sections were dried for 1 h at 65°C before pre-
press the luciferase reporter gene, as previously de-
treatment procedure of deparaffinization, rehydration
scribed [36]. Stable infected cells expressing luciferase
and epitope retrieval in the Pre-Treatment Module,
were selected with Blasticidine S HCl (3 μg/ml; Invitro-
PT-LINK (DAKO) at 95°C for 20 min in Citrate buffer
gen, CA, USA).
(10 mM), Low pH, endogenous peroxidase was blocked
before staining with antibodies against Epithelial Related
Antigen (clone MOC-31, dil. 1:50; DAKO, Denmark). RNA extraction and real-time PCR
After incubation, the reaction was visualized with the Paired frozen tumors and metastatic lymph node samples
EnVision FLEX Detection Kit (DAKO) using diamino- were disrupted and homogenized using a TissueLyser II
benzidine chromogen as a substrate. (Qiagen, Valencia, CA, USA) and total RNA was extracted
Positivity ranged from 75% to 100% with a mean of using the AllPrep DNA/RNA/Protein Mini Kit (Qiagen,
93%. Valencia, CA, USA) according to the manufacturer’s
protocol. Alternatively, total RNA was isolated from
CTC immunoisolation and quantitative real-time Hec1A and Hec1A-ETV5 cells using High Pure RNA
polymerase chain reaction (RT-qPCR) Isolation Kit (Roche Applied Science, Indianapolis, IN,
CTC immunoisolation with the CELLection™ Epithelial En- USA) according to the manufacturer.
rich kit (Invitrogen, Dynal, Oslo, Norway), RNA extraction cDNA synthesis was carried out using MuLV reverse
and RT-qPCR were carried out as previously described [7]. transcriptase Kit (Applied Biosystems, Foster City, CA,
After EpCAM-based immune-enrichment of CTC follo- USA) following the instructions provided by the manu-
wing manufacturer’s protocol, RNA purification was per- facturer. Real-time PCR was performed using Applied
formed with QiampViral kit (Qiagen, Valencia, CA, USA), Biosystem 7500 Real-Time PCR Machine and data
optimized for very low cellularity samples. cDNA was were analyzed with StepOne Software v.2.1 (Applied
synthesized using SuperScriptIII chemistry (Invitrogen, Biosystems, Foster City, CA, USA). GAPDH was used
Carlsbad, CA, USA) according to the user’s guide and as an internal normalization control. The results were
Alonso-Alconada et al. Molecular Cancer 2014, 13:223 Page 9 of 10
http://www.molecular-cancer.com/content/13/1/223

represented as fold change in gene expression relative Abbreviations

to GAPDH gene expression (2-ΔΔCt). EC: Endometrial cancer; CTC: Circulating tumor cells; ENITEC: European
Network for Individualized Treatment in Endometrial Cancer; EMT: Epithelial-
to-mesenchymal transition; FDA: Food and Drug Administration.
In vivo assay and bioluminescent imaging
Six Five-week-old female athymic Nude-Foxn1nu mice Competing interests
The authors declare that they have no competing interest.
were purchased from Harlan Laboratories (Indianapolis,
IN). Mice were divided in 2 groups and either Hec1A or Authors’ contributions
Hec1A-ETV5 stably expressing luciferase cells (5×105 cells LAA, LMR, KM, ADL, CK, JM, EC, JC and XD carried out the immunoisolation,
in 100 μl of sterile PBS) were inoculated into animals by processing and analysis of samples; EW, DH, LC, AC, participated in the
design of the study and critical revision of the manuscript; JT, AGM, LC, JC,
intracardiac injection under 2% isoflurane/air anesthesia. MV, EO and MC contributed to the design and organization of human
A successful intracardiac injection was indicated on day 0 sample collection; HWN, TB, JSG, AR, JR, RLL, HBS, FA, XMG, GMB and MA
by images showing systemic bioluminescence distributed made substantial contributions to conception and design of the study,
interpretation of data and drafting and critical revision of the manuscript. All
throughout the animal. Only mice with evidence of a satis- authors read and approved the final manuscript.
factory injection continued in the experiment. Three
weeks after cells injection and before sacrifice, IVIS system Acknowledgements
We thank the patients for their willingness to participate in the study.
(Xenogen Corporation) coupled to Living Imaging soft- Antonio Diaz-Lopez is a Sara Borrell Fellowship; Laura Muinelo-Romay is
ware 4.2 (Xenogen Corporation) were used to detect supported by ISCIII as Responsible of the Liquid Biopsy Analysis Unit (IDIS);
tumor metastases by bioluminescent imaging. For non- Javier Mariscal is a fellowship from Escola de Doutoramento Internacional
Campus Vida. IISC RETIC- RD12/0036/0007 and RD12/0036/0035.
invasive bioluminescence tumour imaging, luciferin
(Firefly Luciferin, Caliper Lifescience Corp, Hopkinton, Financial support
MA, USA) was used as the substrate for the luciferase ISCIII PI11/00873; Fundación Asociación Española Contra el Cancer (AECC),
Grupos Estables 2011; InveNNta (Innovation in Nanomedicine), co-financed
expressing tumor cells and injected intraperitoneally at a by the European Union (EU) through the Operational Programme for
concentration of 150 mg/kg in PBS. Mice were housed Cross-border Cooperation, Spain-Portugal (POCTEP 2007-2013), European
and maintained under specific pathogen-free conditions Regional Development Fund (ERDF); Helse Vest, Research Council of
Norway, Norwegian Cancer Society and Harald Andersens legat (H.B.S.);
and used in accordance with institutional guidelines ap- L. Alonso-Alconada is recipient of fellowship from the Basque Government
proved by the Use Committee for Animal Care. (Spain).

Author details
Statistical analysis 1
Translational Medical Oncology; Health Research Institute of Santiago (IDIS),
Statistical analyses were conducted using SPSS (Chicago, SERGAS, Trav. Choupana s/n 15706, Santiago de Compostela, Spain.
2
version 15.00 for Windows) and GraphPad Prism 4.00 Department of Obstetrics and Gynecology, Haukeland University Hospital,
Bergen, Norway. 3Department of Clinical Science, Centre for Cancer
software (GraphPad Softwares Inc, San Diego, CA, USA). Biomarkers, University of Bergen, Bergen, Norway. 4Departament of
Mann–Whitney and Kruskal-Wallis non-parametric tests Biochemistry, Universidad Autónoma de Madrid (UAM), Instituto de
were used to determine the differences between con- Investigaciones Biomédicas “Alberto Sols” (CSIC-UAM), IdiPAZ, Madrid, Spain.
5
Department of Pathology, Haukeland University Hospital, Bergen, Norway.
ditions. For Kruskal-Wallis analysis we used Dunn’s post- 6
Department of Women’s and Children’s Health, Institute of Translational
test. Alternatively, Wilcoxon signed test was used to Medicine, University of Liverpool, Liverpool Women’s Hospital, Crown Street,
determine the differences in relative gene expression bet- Liverpool, UK. 7Division of Gynecologic Oncology, Department of Oncology,
KU Leuven, B-3000 Leuven, Belgium. 8Research Unit in Biomedicine and
ween paired frozen tumors and metastatic lymph node Translational and Pediatric Oncology, Vall d’Hebron Research Institute and
samples. Statistical significance was set at p < 0.05. Hospital and Universitat Autònoma de Barcelona, Barcelona, Spain. 9Hospital
MD Anderson Cancer Centre Madrid, Madrid, Spain. 10Department of
Pathology and Molecular Genetics and Research Laboratory, Hospital
Additional files Universitari Arnau de Vilanova, University of Lleida, IRBLLEIDA, Lleida, Spain.
11
Clinical and Surgical Gynaecology R&D + I, Institut Universitari Dexeus,
Additional file 1: CTC analysis using CellSearch technology. Barcelona, Spain. 12University Medical Center Groningen, Department of
Gynecologic Oncology, University of Groningen, P.O. Box 30.001, 9700 RB
Additional file 2: Representative references and corresponding
Groningen, The Netherlands. 13Department of Pathology, Leiden University
RTqPCR Taqman probes of genes covering the main biological
Medical Center, Leiden, The Netherlands. 14GROW: School for Oncology and
functions assessed in CTC immunoisolated from high-risk EC
Developmental Biology, Department of Obstetrics and Gynaecology,
patients.
Maastricht University Medical Centre, Maastricht, The Netherlands.
Additional file 3: Satistical p-values corresponding to the correlation 15
Departament de Ciencies Basiques, Universitat Internacional de Catalunya,
of CTC-gene expression with clinical and pathologic parameters of Barcelona, Spain.
high-risk EC patients included in the study.
Additional file 4: ZEB2 expression in paired samples of primary Received: 13 February 2014 Accepted: 19 September 2014
carcinoma (white box) and affected lymph nodes (grey box) from 6 Published: 27 September 2014
EC patients. ZEB2 increased expression in lymph node metastasis
compared to primary lesions further reinforced the EMT phenotype in References
EC CTC. 1. Yeramian A, Moreno-Bueno G, Dolcet X, Catasus L, Abal M, Colas E,
Additional file 5: EpCAM expression assessed in Hec1A and Reventos J, Palacios J, Prat J, Matias-Guiu X: Endometrial carcinoma:
Hec1A-ETV5 cells using flow cytometry. molecular alterations involved in tumor development and progression.
Oncogene 2013, 32:403–413.
Alonso-Alconada et al. Molecular Cancer 2014, 13:223 Page 10 of 10
http://www.molecular-cancer.com/content/13/1/223

2. Paterlini-Brechot P: Organ-specific markers in circulating tumor cell 22. Kang Y, Massague J: Epithelial-mesenchymal transitions: twist in
screening: an early indicator of metastasis-capable malignancy. Future development and metastasis. Cell 2004, 118:277–279.
Oncol 2011, 7:849–871. 23. Thiery JP, Morgan M: BREAST cancer progression with a TWIST. Nat Med
3. Alix-Panabieres C, Pantel K: Circulating tumor cells: liquid biopsy of 2004, 10:777–778.
cancer. Clin Chem 2013, 59:110–118. 24. Larue L, Bellacosa A: Epithelial-mesenchymal transition in development
4. Negin BP, Cohen SJ: Circulating tumor cells in colorectal cancer: past, and cancer: role of phosphatidylinositol 3′ kinase/AKT pathways.
present, and future challenges. Curr Treat Options Oncol 2010, 11:1–13. Oncogene 2005, 24:7443–7454.
5. Costa C, Abal M, Lopez-Lopez R, Muinelo-Romay L: Biosensors for the 25. Wik E, Raeder MB, Krakstad C, Trovik J, Birkeland E, Hoivik EA, Mjos S,
detection of circulating tumour cells. Sensors (Basel) 2014, 14:4856–4875. Werner HM, Mannelqvist M, Stefansson IM, Oyan AM, Kalland KH, Akslen LA,
6. Barbazan J, Alonso-Alconada L, Muinelo-Romay L, Vieito M, Abalo A, Salvesen HB: Lack of estrogen receptor-alpha is associated with
Alonso-Nocelo M, Candamio S, Gallardo E, Fernandez B, Abdulkader I, de epithelial-mesenchymal transition and PI3K alterations in endometrial
Los Ángeles Casares M, Gómez-Tato A, López-López R, Abal M: Molecular carcinoma. Clin Cancer Res 2013, 19:1094–1105.
characterization of circulating tumor cells in human metastatic colorectal 26. Wicha MS, Liu S, Dontu G: Cancer stem cells: an old idea–a paradigm
cancer. PLoS One 2012, 7:e40476. shift. Cancer Res 2006, 66:1883–1890. discussion 1895–1886.
7. Barbazan J, Vieito M, Abalo A, Alonso-Alconada L, Muinelo-Romay L, 27. Aktas B, Tewes M, Fehm T, Hauch S, Kimmig R, Kasimir-Bauer S: Stem
Alonso-Nocelo M, Leon L, Candamio S, Gallardo E, Anido U, Doll A, cell and epithelial-mesenchymal transition markers are frequently
De los Ángeles Casares M, Gómez-Tato A, Abal M, López-López R: A logistic overexpressed in circulating tumor cells of metastatic breast cancer
model for the detection of circulating tumour cells in human metastatic patients. Breast Cancer Res 2009, 11:R46.
colorectal cancer. J Cell Mol Med 2012, 16:2342–2349. 28. Theodoropoulos PA, Polioudaki H, Agelaki S, Kallergi G, Saridaki Z,
8. Cohen SJ, Punt CJ, Iannotti N, Saidman BH, Sabbath KD, Gabrail NY, Picus J, Mavroudis D, Georgoulias V: Circulating tumor cells with a putative stem
Morse M, Mitchell E, Miller MC, Doyle GV, Tissing H, Terstappen LW, Meropol cell phenotype in peripheral blood of patients with breast cancer. Cancer
NJ: Relationship of circulating tumor cells to tumor response, Lett 2010, 288:99–106.
progression-free survival, and overall survival in patients with metastatic 29. Peacock CD, Watkins DN: Cancer stem cells and the ontogeny of lung
colorectal cancer. J Clin Oncol 2008, 26:3213–3221. cancer. J Clin Oncol 2008, 26:2883–2889.
9. Cristofanilli M, Budd GT, Ellis MJ, Stopeck A, Matera J, Miller MC, Reuben JM, 30. Mirantes C, Espinosa I, Ferrer I, Dolcet X, Prat J, Matias-Guiu X: Epithelial-to-
Doyle GV, Allard WJ, Terstappen LW, Hayes DF: Circulating tumor cells, mesenchymal transition and stem cells in endometrial cancer. Hum
disease progression, and survival in metastatic breast cancer. N Engl J Pathol 2013, 44:1973–1981.
Med 2004, 351:781–791. 31. Castilla MA, Moreno-Bueno G, Romero-Perez L, Van De Vijver K, Biscuola M,
10. de Bono JS, Scher HI, Montgomery RB, Parker C, Miller MC, Tissing H, Doyle Lopez-Garcia MA, Prat J, Matias-Guiu X, Cano A, Oliva E, Palacios J: Micro-
GV, Terstappen LW, Pienta KJ, Raghavan D: Circulating tumor cells predict RNA signature of the epithelial-mesenchymal transition in endometrial
survival benefit from treatment in metastatic castration-resistant carcinosarcoma. J Pathol 2011, 223:72–80.
prostate cancer. Clin Cancer Res 2008, 14:6302–6309. 32. Brabletz S, Brabletz T: The ZEB/miR-200 feedback loop–a motor of cellular
11. Colas E, Muinelo-Romay L, Alonso-Alconada L, Llaurado M, Monge M, plasticity in development and cancer? EMBO Rep 2010, 11:670–677.
Barbazan J, Gonzalez M, Schoumacher M, Pedrola N, Ertekin T, Devis L, 33. Wang G, Guo X, Hong W, Liu Q, Wei T, Lu C, Gao L, Ye D, Zhou Y, Chen J,
Ruiz A, Castellvi J, Doll A, Gil-Moreno A, Vazquez-Levin M, Lapyckyj L, Wang J, Wu M, Liu H, Kang J: Critical regulation of miR-200/ZEB2 pathway
Lopez-Lopez R, Robine S, Friederich E, Castro M, Reventos J, Vignjevic D, in Oct4/Sox2-induced mesenchymal-to-epithelial transition and induced
Abal M: ETV5 cooperates with LPP as a sensor of extracellular signals and pluripotent stem cell generation. Proc Natl Acad Sci U S A 2013,
promotes EMT in endometrial carcinomas. Oncogene 2012, 31:4778–4788. 110:2858–2863.
12. Monge M, Colas E, Doll A, Gonzalez M, Gil-Moreno A, Planaguma J, Quiles 34. Kang Y, Pantel K: Tumor cell dissemination: emerging biological insights
M, Arbos MA, Garcia A, Castellvi J, Llaurado M, Rigau M, Alazzouzi H, from animal models and cancer patients. Cancer Cell 2013, 23:573–581.
Xercavins J, Alameda F, Reventos J, Abal M: ERM/ETV5 up-regulation plays 35. Chambers AF, Groom AC, MacDonald IC: Dissemination and growth of
a role during myometrial infiltration through matrix metalloproteinase-2 cancer cells in metastatic sites. Nat Rev Cancer 2002, 2:563–572.
activation in endometrial cancer. Cancer Res 2007, 67:6753–6759. 36. Sorolla A, Yeramian A, Valls J, Dolcet X, Bergada L, Llombart-Cussac A,
13. Salvesen HB, Haldorsen IS, Trovik J: Markers for individualised therapy in Marti RM, Matias-Guiu X: Blockade of NFkappaB activity by Sunitinib
endometrial carcinoma. Lancet Oncol 2012, 13:e353–e361. increases cell death in Bortezomib-treated endometrial carcinoma cells.
14. Ma PC, Blaszkowsky L, Bharti A, Ladanyi A, Kraeft SK, Bruno A, Skarin AT, Mol Oncol 2012, 6:530–541.
Chen LB, Salgia R: Circulating tumor cells and serum tumor biomarkers in
small cell lung cancer. Anticancer Res 2003, 23:49–62. doi:10.1186/1476-4598-13-223
15. Obermayr E, Sanchez-Cabo F, Tea MK, Singer CF, Krainer M, Fischer MB, Cite this article as: Alonso-Alconada et al.: Molecular profiling of
Sehouli J, Reinthaller A, Horvat R, Heinze G, Tong D, Zeillinger R: circulating tumor cells links plasticity to the metastatic process in
endometrial cancer. Molecular Cancer 2014 13:223.
Assessment of a six gene panel for the molecular detection of
circulating tumor cells in the blood of female cancer patients.
BMC Cancer 2010, 10:666.
16. Taylor SE, Green JA: Towards individualisation of treatment in
endometrial cancer. Br J Cancer 2012, 106:1581–1582.
17. Mariani A, Dowdy SC, Keeney GL, Long HJ, Lesnick TG, Podratz KC: High-risk
endometrial cancer subgroups: candidates for target-based adjuvant
therapy. Gynecol Oncol 2004, 95:120–126. Submit your next manuscript to BioMed Central
18. Lepine J, Audet-Walsh E, Gregoire J, Tetu B, Plante M, Menard V, Ayotte P, and take full advantage of:
Brisson J, Caron P, Villeneuve L, Bélanger A, Guillemette C: Circulating
estrogens in endometrial cancer cases and their relationship with tissular
• Convenient online submission
expression of key estrogen biosynthesis and metabolic pathways. J Clin
Endocrinol Metab 2010, 95:2689–2698. • Thorough peer review
19. Bosse T, Ter Haar NT, Seeber LM, Diest PJ, Hes FJ, Vasen HF, Nout RA, • No space constraints or color figure charges
Creutzberg CL, Morreau H, Smit VT: Loss of ARID1A expression and its
relationship with PI3K-Akt pathway alterations, TP53 and microsatellite • Immediate publication on acceptance
instability in endometrial cancer. Mod Pathol 2013, 26:1525–1535. • Inclusion in PubMed, CAS, Scopus and Google Scholar
20. Salvesen HB, Werner HM, Krakstad C: PI3K pathway in gynecologic
• Research which is freely available for redistribution
malignancies. Am Soc Clin Oncol Educ Book 2013, 218–221.
21. Slomovitz BM, Coleman RL: The PI3K/AKT/mTOR pathway as a therapeutic
target in endometrial cancer. Clin Cancer Res 2012, 18:5856–5864. Submit your manuscript at
www.biomedcentral.com/submit