Published OnlineFirst March 1, 2011; DOI: 10.1158/1535-7163.

MCT-10-0917

Molecular
Cancer
Preclinical Development Therapeutics

A Human Model of Epithelial to Mesenchymal Transition to

Monitor Drug Efficacy in Hepatocellular Carcinoma Progression
Franziska van Zijl1, Sabine Mall1, Georg Machat1, Christine Pirker1, Robert Zeillinger2,
Andreas Weinhaeusel3, Martin Bilban4, Walter Berger1, and Wolfgang Mikulits1

Abstract
The epithelial to mesenchymal transition (EMT) of malignant hepatocytes is a crucial event in hepatocel-
lular carcinoma (HCC) progression and recurrence. We aimed to establish a human model of EMT to examine
drug efficacy and specificity in HCC progression. Human HCC cell populations were characterized by
immunofluorescence analysis, migration and invasion assays, array comparative genomic hybridization,
whole-genome expression profiling, and promoter methylation. Therapeutic agents clinically used against
HCC were examined for efficacy by determination of IC50 values. We show that liver cancer cell lines
exhibited either an epithelial or mesenchymal phenotype of which the latter showed strong migratory and
invasive abilities in vitro. The common cellular origin of both cell types indicated that mesenchymal HCC cells
have been derived from epithelial hepatocytes through EMT in the HCC patient. Drug exposure of
mesenchymal HCC cells showed higher resistance to the targeted therapeutic agents sorafenib and erlotinib
as compared to epithelial HCC cells, which were slightly more resistant to cytostatic drugs. Most remarkably,
combined treatment with doxorubicin and sorafenib caused increased susceptibility of both HCC cell types
resulting in enhanced drug efficacy. Taken together, this EMT model of human HCC allows the identification
of molecular mechanisms and the assessment of therapeutic drug efficacy during liver cancer progression in
preclinical studies. Mol Cancer Ther; 10(5); 850–60.
2011 AACR.

Introduction NF-kB-dependent pathways can induce and maintain

EMT during cancer progression (1, 3).
The epithelial to mesenchymal transition (EMT) enables The pivotal role of EMT in hepatocellular carcinoma
carcinoma cells to invade into surrounding tissues and to (HCC) has been increasingly recognized and various
form secondary tumors known as metastases. A particular molecular mechanisms of hepatocellular EMT have been
characteristic of EMT is the downregulation of E-cadherin identified (4, 5). In human HCC, EMT correlates with
expression, which causes disruption of cell–cell junctions invasive tumors, intrahepatic metastasis, and poor survi-
and dissemination of cells from the primary tumor (1). val (6). This is of particular relevance because intrahepatic
Dysregulation of E-cadherin is provoked by its transcrip- metastases were observed in more than 30% of HCC cases
tional repressors involving Snail/SNAI1, Slug/SNAI2, after liver surgery and in 80% of HCC autopsy cases (7).
ZEB1/DEF1, ZEB2/SIP1, or Twist (2). Receptor tyrosine Epidemiologically, HCC has a poor prognosis and
kinase/Ras and transforming growth factor (TGF)-b sig- represents the third most cause of death from cancer
naling as well as Wnt/b-catenin-, Notch-, Hedgehog-, and worldwide due to diagnosis at advanced stages and lack
of effective therapy options (8). Thus, HCC therapy is
hampered by the fact that the liver is central for xeno-
Authors' Affiliations: Departments of 1Medicine I, Division: Institute of biotic metabolism resulting in rapid modifications and
Cancer Research, Comprehensive Cancer Center, 2Obstetrics and Gyne-
cology, Ludwig Boltzmann Institute for Gynecology and Gynecological efflux of drugs. Curative therapies such as resection and
Oncology, Medical University of Vienna, 3Austrian Institute of Technology, liver transplantation are applicable in only 15% of HCC
and 4Clinical Institute for Medical and Chemical Laboratory Diagnostics, patients (9) and show a high incidence of recurrence (10).
Medical University of Vienna, Vienna, Austria
Unresectable HCC are treated with locoregional thera-
Note: Supplementary data for this article are available at Molecular Cancer
Therapeutics (http://mct.aacrjournals.org/). pies involving transarterial chemoembolization (TACE),
percutaneous ethanol injection, or radiofrequency abla-
F. van Zijl and S. Mall contributed equally to this work.
tion (11). TACE represents an intraarterial administration
Corresponding Author: Wolfgang Mikulits, Department of Medicine I,
Division: Institute of Cancer Research, Medical University of Vienna, of therapeutic drugs combined with embolizing agents
Borschkegasse 8a, 1090 Vienna, Austria. Phone: 43-1-4277-65250; which leads to a more selective distribution and a higher
Fax: 43-1-4277-65239; E-mail: wolfgang.mikulits@meduniwien.ac.at retention time of therapeutics within HCC. TACE has
doi: 10.1158/1535-7163.MCT-10-0917 been established as the standard therapy for patients with

2011 American Association for Cancer Research. intermediate stage cancer (12), although the efficacy of

850 Mol Cancer Ther; 10(5) May 2011

Downloaded from mct.aacrjournals.org on November 1, 2021. © 2011 American Association for Cancer
Research.
 Published OnlineFirst March 1, 2011; DOI: 10.1158/1535-7163.MCT-10-0917

Monitoring Drug Efficacy in Hepatocellular Carcinoma

TACE is limited and patients often suffer from recurrence and 3sp cells, passage numbers between 40 and 50 were
(13). The most frequently used chemotherapeutics for used. All cells were kept at 5% CO2 and 37
C and
TACE are doxorubicin (Adriamycin RDF), cisplatin (Pla- routinely screened for the absence of mycoplasma.
tinol), and epirubicin (Ellence), either alone or in combi-
nation (14), from which doxorubicin and cisplatin Analysis of diploidy
showed significant benefits (13). Moreover, targeted ther- Cells were trypsinized, washed with PBS and 1  106
apeutic agents such as the multikinase inhibitor sorafenib cells were incubated with 1 mL hypotonic DNA stain-
(Nexavar) have been shown to have survival benefits ing buffer containing 5 mg/mL propidium iodide
especially in patients with advanced stage HCC (15– (Calbiochem), 0.1% sodium citrate, 0.3% Triton X-100
17). Those studies are the basis to develop treatment and 20 mg/mL ribonuclease A (all from Sigma) for 30
modalities to efficiently combat HCC progression that minutes on 4
C in the dark. Determination of the DNA
is considered as utmost treatment-resistant. However, content was done by flow cytometry (FACSCalibur, BD
suitable human HCC models are required to identify Biosciences).
potential molecular targets, to test drug efficacy and to
estimate the effective concentration of therapeutic agents. Short tandem repeat analysis
Here, we established a human model of hepatocellular Polymerase chain reaction (PCR) of 7 short tandem
EMT which reflects important aspects of HCC progres- repeats (Supplementary Table S1) was done to verify the
sion and allows studying the underlying molecular genetic origin of human 3p and 3sp hepatoma cells.
mechanisms. The matched pair of epithelial and
mesenchymal HCC cells enables us (i) to evaluate the Quantitative real time PCR
efficacy of currently used therapeutic agents in single or PCR reactions were done with Fast SYBR Green
combined treatments and (ii) to assess the effectiveness of Master Mix (Applied Biosystems) in duplicates according
novel anticancer agents during HCC progression. to the recommendations of the manufacturer and quan-
tified with the 7500 Fast Real Time PCR System (Applied
Material and Methods Biosystems). Arbitrary units were calculated by the dCT
method. Forward and reverse primers are described in
Cell culture Supplementary Table S1.
HCC-1.2 and HCC-1.1 cells, referred to as 3p and 3sp
cells, respectively, were established from 1 male patient Confocal immunofluorescence microscopy
diagnosed with HCC (grade 2, T2N0M0, hepatitis C virus Cells were seeded onto SuperFrostPlus glass slides
positive) at the age of 56 years. A fully documented (Menzel) and fixed with either 4% paraformaldehyde
patient’s history and informed consent was obtained. or methanol/aceton. Collagen gels were fixed in 4%
The study protocol conforms to the ethical guidelines formaldehyde/PBS for 10 minutes at room temperature.
of the 1975 Declaration of Helsinki, as reflected by the After blocking for 30 minutes (1% PBS/0.5% Tween 20/
approval of the Ethic Committee of the Vienna Medical 0.2% fish gelatine with 0.2 mg/mL IgG1) at room tem-
University. Briefly, the primary liver tumor tissue from perature, the following primary antibodies were applied
surgical resection of the HCC patient was cut into pieces at a dilution of 1:100: anti-b-catenin (BD Transduction
of approximately 0.5 mm3 and seeded on collagen I Laboratories), anti-E-cadherin (BD Transduction Labora-
(Sigma) coated tissue culture plates by incubation in tories), anti-p120catenin (p120ctn; BD Transduction
ACL4 medium plus 5% fetal calf serum (FCS) at 37
C Laboratories), anti-keratin 8 (Progen), Texas Red-X phal-
and 5% CO2. Outgrowing epithelial 3p and fibroblastoid loidin (Invitrogen), and anti-vimentin (Sigma). The cor-
3sp cells were further separately cultivated. The propa- responding secondary antibody (Alexa 488; Invitrogen)
gation and immortalization of epithelial 3p and fibro- was applied after 1 hour. Nuclei were visualized with To-
blastoid 3sp cell populations were done by passaging Pro3 (Invitrogen) at a dilution of 1:10,000. Imaging was
cells at a ratio of 1:2 once a week in ACL-4 medium done by confocal immunofluorescence microscopy
supplemented with 5% FCS. Fibroblasts were eliminated (Zeiss).
from the epithelial 3p cell culture between 12 and
14 weeks by differential trypsinization over 2–3 cell Proliferation analysis
passages. The established cell lines that have undergone The 1.5  104 cells were plated and cell numbers were
more than 30 cell doublings were maintained in RPMI determined after various time points with a multichannel
1640 and 10% FCS. The 3p hepatoma cells were further cell analyzer (CASY; Sch€ arfe Systems). Three indepen-
cultivated on collagen I coated tissue culture plates dent experiments were done in triplicates.
whereas 3sp cells were propagated on tissue culture
plastic. 3p cells of passages 10 to 12 and 3sp cells of Cell migration and invasion assays
passages 7 to 13 are termed 3p early and 3sp early, Cell migration and cell invasion was determined by
respectively. 3p cells between passage 71 and 76 and Platypus Technology according to the manufacturer’s
3sp from passage 72 to 87 were termed 3p late and 3sp protocol (Oris Cell Invasion & Detection Assay). To
late, respectively. Without particular designation of 3p analyze cell migration, 5  104 cells were plated onto 8

www.aacrjournals.org Mol Cancer Ther; 10(5) May 2011 851

Downloaded from mct.aacrjournals.org on November 1, 2021. © 2011 American Association for Cancer
Research.
 Published OnlineFirst March 1, 2011; DOI: 10.1158/1535-7163.MCT-10-0917

van Zijl et al.

not-coated wells. To examine cell invasion, same cell CATGCGGAAT-30 ; exon 9, for 50 -CCTGACTCATAG-
numbers were seeded into 8 basal membrane extract CAAGGGCCCCA-30 , rev 50 -GCTGTCAGGATAGCCAA-
(BME; 3.5 mg/mL)-coated wells and overlaid with undi- ATCAATGTCC-30 ; exon 10, for 50 -ATGCCTCGGCC-CCA-
luted BME. Migration and invasion of cells was micro- GACCAA-30 , rev 50 -AGATGGCGCCTGCAGAGGGA-30 ;
scopically analyzed and quantitatively evaluated by exon 12, for 50 -TGAACTGCCCTCCACCAGCCA-30 , rev
measuring the fluorescence after staining with Cell- 50 -GGTCAGTCAGGCAGCTCCCCA-30 ; exon 15, for 50 -
Tracker (Green CMFDA; Invitrogen). AGGACAGGAGCTGGCCTCGT-30 , rev 50 -CCTGTGTG-
TGCATCACTGGGGT-30 ; exon 18, for 50 -GCCTGGCC-
3-Dimensional spheroid formation of HCC cells TTCCCAGGCTAC-30 , rev 50 -TGGACGTTGGCTCCTTG-
Spheroid formation and incubation was done as TGGC-30 ; exon 21, for 50 -TCCCTCTGCTTCCCCACCCT-
described (18). Briefly, a cell suspension of 3  102 3p 30 , rev 50 -TCAACACATGGCCCTGGGA-GT-30 ; exon 24,
cells per 100 mL or a 1:1 mixture of 3p and 3sp cells in for 50 -TCTCTGCCACACCCCCAGCC-30 , rev 50 -TCTG-
RPMI containing 20% methyl cellulose (Sigma) was incu- CAGGAGGCAGGTGTCCA-30 ; exon 25, for 50 -
bated for 3 days at 37
C and 5% CO2. After harvesting, 96 CTCCTGCAGCCCCTGCCAAC-30 , rev 50 -GGGAAGG-
spheroids were mounted in collagen I (Sigma) and plated CAGCTCTGTCCGC-30 . Human AXIN1: exon 1, for 50 -
into flexiPERM conA wells (Greiner Bio-One GmbH) to CCGGCTACCCGAGATACTCAGCA-30 , rev 50 -GCTCCA-
harden at 37
C for 15 minutes. Gels were then transferred GGCATCATGGCAGCAA-30 ; exon 2, for 50 -CCCCCAG-
into a 24-well plate and incubated with medium at cell TGCCTGGTGAGGA-30 , rev 50 -TGGGGCTCTCCGAGC-
culture conditions as outlined. CellTracker (Green CTGTC-30 ; exon 3, for 50 -TCCTGGCGGGAGCCAGT-
CMFDA; Invitrogen) was applied at a concentration of CAA-30 , rev 50 -CTGCACCAGCGTCGGCAGAA-30 ; exon
5 mM before spheroid formation and incubated as out- 4, for 50 -ACAGGCAGGCACAAGGAGGC-30 , rev 50 -
lined by the instructions of the manufacturer (Invitro- CCCTGCTGCCCTCAGGGACT-30 ; exon 5, for 50 -
gen). The fluorescence signal was visualized by GGCCGGACACGAAAAGGGGG-30 , rev 50 -CATCAGGA-
microscopy (Nikon, TE 300). CATCCCGGCGGC-30 ; exon 6, for 50 -CGGCGATCCATC-
GTCAGGGC-30 , rev 50 -GCGAGGCCATCACTGGCGTT-
Array comparative genomic hybridization 30 ; exon 7, for 50 -AGGATGCGGAGAAGAACCAG-30 ,
Genomic DNA was isolated from early and late 3p and rev 50 -AGAAGTGCTCACGCTACACA-30 ; exon 8, for 50 -
3sp HCC cells using the QIAamp DNA Blood Mini Kit ACGAGGAAGCCACAGCCCCA-30 , rev 50 -TGCCTCCA-
(Qiagen) according to manufacturer’s protocol. Array GGACCTCTGCCC-30 ; exon 9, for 50 -AGACCGAGTTGC-
comparative genomic hybridization (aCGH) analysis CACGCACG-30 , rev 50 -CCGGGAGGACCCTCAGGACG-
was done using human 4  44K whole genome oligonu- 30 ; exon 10, for 50 -TGAGGCCAGCAGGAGGGACC-30 , rev
cleotide-based arrays (Agilent). Male reference DNA 50 -GGGCCGAAAGCCTGATGCCA-30 ; exon 11, for 50 -
(Promega) and sample DNA were digested with AluI GGCCGTCCTGCCCGTCTTTG-30 , rev 50 -GGCTGGAGG-
and RsaI (Promega) and labeled by random priming with CAGGTGCAGTG-30 .
Cyanine 3- and Cyanine 5-dUTP (Perkin-Elmer), respec-
tively, by using the BioPrime Array CGH Genomic Label- Expression profiling using microarrays
ing Kit (Invitrogen). After purification with Microcon Total RNA was isolated from duplicates of early and
YM-30 filter units (Millipore), the 2 labeled DNA species late 3p and 3sp HCC cells using the RNeasy Mini kit
were hybridized together with human cot-DNA (Roche) (Qiagen). The integrity and quantity of RNA was ana-
onto CGH arrays for 48 hours at 65
C. Slides were lyzed by Agilent Bioanalyzer (Agilent Technologies).
scanned with a G2505B Micro Array Scanner (Agilent). cDNA labeling and hybridization on Affymetrix Gene-
Feature extraction and data analysis were done using the Chip Human gene 1.0 ST Array (Affymetrix) as well
Feature Extraction and DNA Analytics software (Agi- as scanning of signal intensities was done according
lent), respectively. to the manufacturer’s protocol. The ratio of regulation
was calculated and a minimum of 3-fold regulation was
Polymerase chain reaction of genomic DNA considered as significant. Pathway analysis was per-
Chromosomal DNA was isolated with GenElute Mam- formed with Gene Set Enrichment Analysis software by
malian Genomic DNA Miniprep Kit (Sigma). PCR was comparing the molecular profile data with existing as
done using chromosomal DNA and illustra PuReTaq well as self-defined gene sets. Complete gene expres-
Ready-To-Go PCR beads (GE Healthcare) and 4 mmol/L sion data have been deposited in National Center for
primer mix of following primers: human TRPM3: exon 1, Biotechnology Information’s Gene Expression Omnibus
forward (for) 50 -CAGGCAAGGCTACTTGACTA-30 , and are accessible by GEO Series accession no.
reverse (rev) 5 -CACCAGCCTCTTGTCTCTGA-30 ; exon
0
GSE26391.
3, for 50 -GGTGGACGGAGAAGGGCAGGA-30 , rev 50 -
GGGGCAGGGGACTCAGAGCA-30 ; exon 6, for 50 -CT- Analysis of CpG methylation
CCTAGCTCCCCGCCACCA-30 , rev 50 -TGCCAGTGGG- Genomic DNA was isolated from 3p and 3sp HCC
AACAACCGCT-30 ; exon 8, for 50 -TCCACAATT- cells grown in quadruplicates and from primary
AGACATGACGCTGCAA-30 , rev 50 -GCTGGCCACC- peripheral blood cells using the QIAamp DNA Blood

852 Mol Cancer Ther; 10(5) May 2011 Molecular Cancer Therapeutics

Downloaded from mct.aacrjournals.org on November 1, 2021. © 2011 American Association for Cancer
Research.
 Published OnlineFirst March 1, 2011; DOI: 10.1158/1535-7163.MCT-10-0917

Monitoring Drug Efficacy in Hepatocellular Carcinoma

Mini Kit (Qiagen). The digestion of 600 ng genomic linear interpolation of data points and are depicted by
DNA with methylation-sensitive restriction enzymes dose–response curves using the software GraphPad
(MSRE) was done. In total, 327 cancer-specific promoter Prism 5.01.
loci were analyzed by PCR. Six hundred nanograms of
genomic DNA of 3p and 3sp HCC cells and from Verification of IC50 values by proliferation
primary peripheral blood cells were digested with 6  104 cells/well were seeded in duplicates on col-
MSRE overnight at 37
C by using a mixture of 6 units lagen-coated 12-well plates and incubated with drugs at a
of each AciI (New England Biolabs), Hin6I (Fermentas). concentration of determined IC50 values for 3 days. Cells
and HpaII (Fermentas). Quantitative polymerase chain were trypsinized and counted with a cell counter (CASY,
reaction (qPCR) was done on MSRE-digested DNA Roche). Untreated cells were used as reference to calcu-
covering the CpG rich area at chr5:151,066,289– late the percentage of proliferation.
151,066,501 (human genome annotation: hg; ref. 19),
including part of the first gene coding exon. This Statistical analysis
sequence region contains 9 MSRE sites when using Data were expressed as means  SD. The statisti-
the mixture of 3 enzymes mentioned earlier. Comple- cal significance of differences was evaluated using
tion of digestion was confirmed using a control PCR an unpaired, nonparametric Student’s t test. Signi-
covering known differentially methylated regions ficant differences between experimental groups were
(DMRs, H19, IGF2, ABL1, PITX2, XIST, and FMR1) as *P < 0.05, **P < 0.01, or ***P < 0.005.
published recently (19). Ten nanograms of MSRE-
digested DNA was amplified in 10 mL PCR volume Results
using 0.3 U HotStarTaq polymerase and buffer con-
taining 1.5 mmol/L MgCl2 (Qiagen) using a 384-well HCC cells show an epithelial phenotype and a
format and the LightCycler LC480 (Roche). Cycling mesenchymal one correlating with enhanced cell
was done with 15 minutes of initial denaturation motility
followed by 50 cycles with each 95
C for 40 seconds, We analyzed various human hepatoma cell lines for
65
C for 40 seconds and 72
C for 80 seconds including a their epithelial and mesenchymal characteristics to estab-
final extension at 72
C for 7 minutes. A melting curve lish a model of HCC progression (Supplementary
analysis was done for confirming specificity of ampli- Table S2). Two distinct liver cell lines were of particular
cons with a Tm of 88.4
C  0.6
C. Cp values were interest as these cell types have been isolated from the
extracted from qPCR-data using the LigthCycler HCC of one patient (20). Phase contrast analysis sug-
LC480 software. Box plots were generated from quad- gested an epithelial cell type, termed 3p, and a mesench-
ruplicate 45-Cp values using R statistical software. ymal cell population, designated 3sp (Fig. 1A). Both HCC
cells showed diploid DNA content (Supplementary
Fig. S1) and short tandem repeat analysis verified their
Therapeutic agents
common genomic identity (Supplementary Table S3).
Doxorubicin hydrochloride (adriamycin hydrochlor-
qRT-PCR analysis revealed that 3p cells express epithelial
ide), cis-diammineplatinum (II) dichloride (cisplatin)
markers such as E-cadherin and keratin 8, whereas 3sp
and epirubicin hydrochloride were purchased from
cells showed a mesenchymal expression signature by
Sigma and dissolved in 0.9% NaCl. Sorafenib (Nexavar;
upregulation of the transcription factors LEF1, SNAI1,
Bayer HealthCare Pharmaceuticals), erlotinib (Tarceva;
SNAI2, and ZEB1 (Fig. 1B). Immunolocalization showed
LC Laboratories), and bevacizumab (Avastin; Roche)
intact adherence junctions by expression of E-cadherin,
were dissolved in dimethylsulfoxide (DMSO). Stock solu-
b-catenin, and p120catenin at cell borders of 3p cells,
tions were diluted in medium to concentrations indicated
whereas 3sp cells failed to show this phenotype
in the text.
(Fig. 1C). In contrast to 3sp, 3p cells displayed cytoplas-
mic distribution of the epithelial marker keratin 8 by
Determination of the inhibitory concentration concomitant absence of the mesenchymal interfilament
Cell viability was determined using the 3-(4,5 component vimentin. In addition, mesenchymal 3sp cells
dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide showed relocalization of actin from the cell membrane to
(MTT) assay. Briefly, cells were seeded in triplicates at stress fibers suggesting enhanced motility.
a density of 6  103 cells per well. After 24 hours, cells We next assessed the migratory and invasive potential
were incubated with drug-containing medium for 3 days. of 3p and 3sp cells. First, we did proliferation kinetics to
Cells were incubated with MTT solution (5 mg/mL; exclude an influence of differential cell doubling rates on
Sigma) and medium was replaced with DMSO after 5 migration and invasion assays. No significant difference
hours. The absorbance was measured at 620 nm by using in proliferation of 3p and 3sp cells could be detected
a microplate reader (Asys HiTech). MTT assays were (Fig. 2A). Yet, a 4-fold increased migratory potential of
repeated 3 times for each drug application and untreated mesenchymal 3sp cells was determined when compared
cells were used as reference. Determination of the inhi- to 3p cells (Fig. 2B and 2C). To analyze invasion, HCC
bitory concentration (IC50) values were obtained by log- cells were embedded into matrigel. Although epithelial

www.aacrjournals.org Mol Cancer Ther; 10(5) May 2011 853

Downloaded from mct.aacrjournals.org on November 1, 2021. © 2011 American Association for Cancer
Research.
 Published OnlineFirst March 1, 2011; DOI: 10.1158/1535-7163.MCT-10-0917

van Zijl et al.

Figure 1. Epithelial and

mesenchymal characteristics of
3p and 3sp HCC cells. A, phase
contrast images display an
epithelial (3p) and a fibroblastoid
phenotype (3sp; bar, 50 mm). B,
qRT-PCR analysis of epithelial
(E-cadherin, keratin 8) and
mesenchymal markers (LEF1,
SNAI1, SNAI2, and ZEB1) in 3p
and 3sp cells. C, confocal
immunofluorescence analysis of
the epithelial markers E-cadherin,
b-catenin, p120catenin, and keratin
8 as well as the mesenchymal
marker vimentin. Localization of
actin at cell membranes and actin
stress fibers as indicated by
phalloidin staining (red). Nuclei
were counterstained with To-Pro3
(blue). Bar, 25 mm. dCT, delta CT
value.

3p cells showed no invasive potential, mesenchymal 3sp mal 3sp cells failed to form spheroids on their own,
cells displayed a more than 5-fold stronger ability to however, these cells attached onto the surface of epithe-
invade into the matrix (Fig. 2D and E). The invasive lial 3p spheroids after cocultivation, and showed strong
potential of mesenchymal 3sp cells was further analyzed invasion into the surrounding matrix (Supplementary
by 3-dimensional spheroid formation in collagen gels Fig. S2A and B). Immunofluorescence analysis of E-cad-
(18). Epithelial 3p cells formed compact and round spher- herin and b-catenin indicated epithelial characteristics of
oids without cell invasion into the surrounding gel (Sup- 3p-derived spheroids, whereas these markers could not
plementary Fig. S2A), even though proliferation was not be detected in invading 3sp cells due to disassembly of
impaired (data not shown). On the contrary, mesenchy- epithelial junctions (Supplementary Fig. S2C). Taken

854 Mol Cancer Ther; 10(5) May 2011 Molecular Cancer Therapeutics

Downloaded from mct.aacrjournals.org on November 1, 2021. © 2011 American Association for Cancer
Research.
 Published OnlineFirst March 1, 2011; DOI: 10.1158/1535-7163.MCT-10-0917

Monitoring Drug Efficacy in Hepatocellular Carcinoma

Figure 2. Mesenchymal 3sp HCC

cells exhibit enhanced migration
and invasion. A, proliferation
kinetics of epithelial 3p and
mesenchymal 3sp cells. B,
migration assay using the
Platypus technology after 3 days
and the quantification of cell
migration by fluorometric analysis.
C, invasion assays done by using
the Platypus technology after 7
days. Top left and bottom left
show the invasion area in the
absence and presence of the
mask, respectively. Quantification
of cell invasion by fluorometric
analysis is shown in the right. Cells
were visualized with green
CellTracker. Bar, 250 mm;
***, P < 0.005.

together, 3p and 3sp cells showed a distinct epithelial and located between exons 12 and 15 and between exons 2 and
mesenchymal phenotype correlating with poor and 3 in TRPM3 and AXIN1, respectively (Supplementary
strong migratory and invasive abilities, respectively. Fig. S4). This homozygous loss of chromosomal regions
indicates an identical cellular origin of 3p and 3sp hepa-
Mesenchymal HCC cells developed through toma cells, and thus shows that 3sp cells have been
undergoing EMT in the patient developed from 3p cells via EMT in the patient.
We further investigated whether 3sp cells have been To verify EMT, we analyzed the 3p/3sp HCC cell lines
derived from epithelial 3p cells through EMT in the HCC by whole-genome expression profiling (Fig. 4A and B).
patient. aCGH analysis showed that changes in the geno- Remarkably, a large cluster of liver-specific genes were
mic DNA of these HCC cell lines were widely identical down-regulated in 3sp cells compared to 3p cells (Sup-
(Supplementary Fig. S3), which were exemplified by the plementary Fig. S5), including CYP450 phase I enzymes,
loss of genomic DNA in the TRPM3 and AXIN1 loci of both alcoholdehydrogenase (ADH) and aldehydedehydrogenase
cell types (Fig. 3A and B). PCR analysis of multiple exons (ALDH) and phase II enzymes such as UDP-glucurono-
could identify the exact chromosomal breaks, which are syl-transferases (UGT) and glutathione-S-tranferase (GST).

www.aacrjournals.org Mol Cancer Ther; 10(5) May 2011 855

Downloaded from mct.aacrjournals.org on November 1, 2021. © 2011 American Association for Cancer
Research.
 Published OnlineFirst March 1, 2011; DOI: 10.1158/1535-7163.MCT-10-0917

van Zijl et al.

(Supplementary Fig. S6). From these data we conclude

that mesenchymal 3sp HCC cells derived from epithelial
3p cells via EMT in the HCC patient.

SPARC activation by CpG demethylation

We next addressed the differences in the methylation
status of gene promoters in 3p and 3sp cells. To exclude
cell culture effects, genomic DNA of 3p/3sp cells from
early as well as late passages was analyzed. Most notably,
SPARC was identified to be epigenetically regulated, in
line with a 128-fold upregulation of the transcript level
(Fig. 4B). Methylation-specific PCR of the SPARC pro-
moter revealed a lowered methylation status in both early
and late passaged mesenchymal 3sp cells (Fig. 4C), which
was verified by qRT-PCR analysis showing a >30-fold
increase of SPARC mRNA in 3sp cells (Fig. 4D). Because
the aCGH profile of the SPARC locus was unaffected,
these data provide first evidence that the upregulation of
SPARC is caused by demethylation during HCC progres-
sion.

The 3p/3sp EMT model is suitable for drug testing

We next used this human HCC model to assess the
efficacy of clinically used anti-cancer agents. Dose–
response relationships and corresponding half maximal
inhibitory concentrations (IC50) of drugs were deter-
mined (Supplementary Table S4) and verified by viability
assays (Supplementary Fig. S7A). Chemotherapeutics
such as doxorubicin, cisplatin, and epirubicin currently
used for TACE were compared to targeted therapeutic
drugs such as the multikinase inhibitor sorafenib, the
EGFR-inhibitor eroltinib, and the VEGF-inhibitor bevaci-
zumab. 3sp cells showed a slightly increased suscept-
Figure 3. Mesenchymal 3sp HCC cells are derived from epithelial 3p ibility against chemotherapeutics when compared to 3p
HCC cells through EMT in vivo. The aCGH analysis revealed a cells (Fig. 5A–C; Supplementary Table S4), which is
homozygous loss of genomic DNA in 3p and 3sp cells. Homozygous loss explained by downregulation of multiple drug resistance
of DNA in the TRPM3 (A) and AXIN1 gene locus (B). Red bars indicate
genes in 3sp cells (Supplementary Fig. S8A). Interest-
exons and blue bars denote the positions of arrayed oligonucleotides
on the chromosome. bps, basepairs. ingly, 3sp cells were more resistant to targeted therapies
including sorafenib and erlotinib (Fig. 5D and E; Supple-
mentary Table S4). Bevacizumab did not show an effect
Importantly, a cluster of epithelial markers involving on neither of the cells (Fig. 5F; Supplementary Table S4).
components of tight junctions (claudins), desmosomes Of particular interest was the higher resistance of 3sp
(desmoplakin), and adherence junctions (E-cadherin) were cells towards sorafenib, as expression analysis showed a
found to be strongly down-regulated in mesenchymal 10-fold upregulation of PDGF-Rb (Fig. 4B). Additional
3sp cells (Fig. 4A). Moreover, EMT-specific genes such as targets of sorafenib including VEGFR-1, -2, -3, and RAF
vimentin, LEF1, and SNAI2 were highly up-regulated in kinases were found unaltered between 3p and 3sp
3sp cells (Fig. 4B), as shown in Fig. 1B and C. In addition, cells (data not shown). As PDGFR-b is described to
expression profiling revealed (i) upregulation of growth play an important role in migration (21), we analyzed
factors and matrix metalloproteinases (MMP) such as con- the migratory potential of 3sp cells on sorafenib treat-
nective tissue growth factor (CTGF), TGF-b2, and MMP2; (ii) ment (Supplementary Fig. 7B and 7C). Interestingly,
upregulation of receptors associated with EMT such as delivery of sorafenib at the IC50 showed a strong reduc-
platelet-derived growth factor (PDGF)-Rb and discoidin tion of migration, whereas control treatment of doxoru-
domain receptor tyrosine kinase (DDR)2; and (iii) increase bicin at the IC50 did not impair migration. Together, these
of EMT-associated proteins such as fibroblast activation data indicate that sorafenib has a lower cytotoxic poten-
protein (FAP), lysyl oxidase, tenascin C and SPARC (secreted tial on 3sp cells, however, strongly represses their migra-
protein, acidic and rich in cysteine). qPCR further con- tory abilities.
firmed the upregulation of PDGF-Rb, TGF-b2, DDR2, In recent clinical efforts, sorafenib and doxorubicin
and FAP in 3sp cells as detected by microarrays analysis were frequently used in combination therapy. Thus,

856 Mol Cancer Ther; 10(5) May 2011 Molecular Cancer Therapeutics

Downloaded from mct.aacrjournals.org on November 1, 2021. © 2011 American Association for Cancer
Research.
 Published OnlineFirst March 1, 2011; DOI: 10.1158/1535-7163.MCT-10-0917

Monitoring Drug Efficacy in Hepatocellular Carcinoma

Figure 4. Global transcriptome

analysis shows regulation of EMT-
specific genes in 3sp HCC cells.
Gene Set Enrichment Analysis of a
whole-genome Affymetrix
GeneChip was done.
Downregulation (blue) of epithelial
markers (A) and upregulation (red)
of EMT regulators (B) in 3sp cells
were determined. The fold
upregulation of mRNA was
calculated and depicted by the
ratio of 3p to 3sp cells (A; 3p/3sp)
and of 3sp to 3p cells (B; 3sp/3p).
C, DNA methylation testing by
qPCR on methylation-sensitive
digestion shows demethylation of
the SPARC promoter in 3sp cells.
Four replicates of each cell type
were analyzed. DNA from male
and female blood (each n ¼ 4)
were used as negative controls.
D, confirmation of SPARC
upregulation in 3sp cells by qRT-
PCR. dCT, delta CT value. ACTA2,
smooth muscle actin; CDH1, E-
cadherin; CLDN, claudin; CTGF,
connective tissue growth factor;
DDR2, discoid domain receptor;
DSP, desmoplakin; EPCAM,
epithelial cellular adhesion
molecule; FAP; fibroblast
activation protein; HNF4A,
hepatocyte nuclear factor 4a;
KRT, keratin; LEF1, lymphoid
enhancing factor 1; LOX, lysyl
oxidase; MMP, matrix
metalloproteinase; PDGF-Rb,
platelet-derived growth factor
receptor b; SNAI2, Slug; SPARC,
secreted protein, acidic, cystein-
rich; TGF-b, transforming growth
factor-b; TNC, tenascin C; VIM,
vimentin.

we investigated the effect of the combined treatment of ingly, we observed a nomalization of the IC50 values
these anticancer drugs in our model of HCC progression. between 3p and 3sp cells after combined treatment in
We therefore treated cells with one drug in increasing both approaches (Fig. 6B and 6D; Supplementary
concentrations, whereas the second drug was kept con- Table S4). Taken together, these results indicate that
stant at its IC50. First, variable amounts of doxorubicin the monotherapy shows considerable differences in the
were applied in combination with the IC50 of sorafenib on susceptibility between HCC cells undergoing EMT. Our
3p and 3sp cells (Fig. 6A). Compared to doxorubicin data provide evidence that the combined use of doxor-
monotherapy, cells showed a drastic susceptibility for ubicin and sorafenib is highly efficient to target both
the combined treatment, leading to a reduction of the IC50 epithelial and mesenchymal HCC cells.
for doxorubicin of 60% and 39% in 3p and 3sp cells,
respectively (Fig. 6B; Supplementary Table S4). The vice Discussion
versa treatment, applying variable amounts of sorafenib
in combination with the IC50 of doxorubicin (Fig. 6C) EMT has been increasingly recognized to play a crucial
showed a similar result, leading to a reduction of the IC50 role in HCC progression by the acquisition of invasive
for sorafenib of 18% and 50% in 3p and 3sp cells, respec- properties. In line with the transdifferentiation of neo-
tively (Fig. 6D; Supplementary Table S4). Most interest- plastic hepatocytes to motile mesenchymal derivatives,

www.aacrjournals.org Mol Cancer Ther; 10(5) May 2011 857

Downloaded from mct.aacrjournals.org on November 1, 2021. © 2011 American Association for Cancer
Research.
 Published OnlineFirst March 1, 2011; DOI: 10.1158/1535-7163.MCT-10-0917

van Zijl et al.

Figure 5. Increased resistance of

mesenchymal HCC cells against
targeted anticancer drugs.
Cytotoxicity of cytostatic agents
(A) and targeted therapeutic drugs
(B) was analyzed by MTT assays
and IC50 values were calculated.
Dose–response relationships of
3p and 3sp cells were compared.
Dotted lines indicate IC50.

HCC is described as a heterogenous tumor at advanced tissue invasion, EMT and metastasis, vascular transmi-
stages showing clonal expansion of genetically distinct gration and angiogenesis, earlier recurrence and shor-
malignant cell populations. Therefore, efficient antican- tened survival time as well as downregulation of
cer therapy depends on targeting cancer cells at all stages glycolysis-regulating and acute-phase response genes.
of differentiation. Aligning the gene expression sets of 3p and 3sp cells
EMT has been suggested as the critical step in tumor revealed a correlation with early and late TGF-b signature
cell dissemination and particularly associates with resis- of more than 70 percent, respectively (data not shown).
tance towards chemotherapy and immunotherapy (1). Moreover, we observed upregulation of TGF-b1 in
Here we established and characterized the first human mesenchymal 3sp cells and a resistance against TGF-b
cellular model of EMT in HCC progression which corre- treatment suggesting an autocrine regulatory TGF-b loop.
lates with the recently established molecular expression Thus, this model is a reliable tool to determine novel key
pattern of early and late TGF-b signatures (22). The late players and to test drug efficacy of new and currently
TGF-b signature of HCC was shown to associate with approved anticancer agents within HCC progression. By

858 Mol Cancer Ther; 10(5) May 2011 Molecular Cancer Therapeutics

Downloaded from mct.aacrjournals.org on November 1, 2021. © 2011 American Association for Cancer
Research.
 Published OnlineFirst March 1, 2011; DOI: 10.1158/1535-7163.MCT-10-0917

Monitoring Drug Efficacy in Hepatocellular Carcinoma

Figure 6. Combined treatment

with sorafenib and doxorubicin
efficiently targets both epithelial
and mesenchymal HCC cells.
Combined application was done
by delivering cells with one drug in
increasing concentrations while
the second drug was kept
constant at the IC50
concentration. Dose–response
relationships for 3p and 3sp cells
on treatment with doxorubicin
alone or in combination with
sorafenib (A; þIC50 sora) and
sorafenib alone or in combination
with doxorubicin (C; þIC50 doxo)
are shown. The resulting change
of IC50 values (arrow, %) of single
to combined treatment are
depicted for doxorubicin (B) and
sorafenib (D). Arrow bars indicate
standard deviation of triplicates
(A, C) or range of calculated
IC50 from at least 3 independent
experiments (B, D).

using this model we show that epithelial cells are more with or without doxorubicin treatment in advanced or
sensitive to the targeted therapeutic agents such as sor- metastatic liver cancer is currently recruiting HCC
afenib and erlotinib, whereas mesenchymal cells show a patients. Clinical trial information of the U.S. National
slightly more efficient susceptibility to chemotherapeutic Institutes of Health is available under the identifier
drugs such as doxorubicin, cisplatin, and epirubicin. This NCT01015833.
might be caused by the downregulation of various multi- Remarkably, our data obtained by using 3p/3sp HCC
ple drug resistance proteins in the 3sp cells (Supplemen- cells confirm these observations and thus underline the
tary Fig. S8A). In the same line, the high sensitivity synergistic effects of sorafenib and doxorubicin in inter-
towards erlotinib might be explained by the higher phos- fering with HCC progression. Other ongoing clinical
phorylation status of EGFR found in 3p cells (Supple- trials investigate the combination of sorafenib and dox-
mentary Fig. S8B). orubicin or other chemotherapeutics using TACE. Clin-
We propose that the molecular changes on EMT ical trial information of the U.S. National Institutes of
are crucial for differential drug efficacy. In this regard, Health is available under the identifiers NCT01011010,
we could show for the first time in vitro that the NCT00478374 and NCT0085528. Interestingly, treatment
combined application of sorafenib and doxorubicin of the 3p/3sp HCC model with both sorafenib and
shows 2 advantages. First, combined therapy is capable erlotinib (currently used in a phase III study) also
of targeting both, epithelial and mesenchymal cells, revealed to be effective at reduced concentrations while
which might have the potential to reduce the risk of targeting both epithelial and mesenchymal HCC cells
HCC recurrence. Second, the effective concentrations (data not shown).
of drugs could be reduced and thus side effects can Our data revealed a significant upregulation of SPARC
be minimized. Yet, both cell lines showed no in mesenchymal 3sp cells which has been reported in
response against bevacizumab, referring to the limita- patient samples at advanced stages of disease (24). The
tion of the HCC model to study angiogenic mechan- role of this protein in hepatocarcinogenesis, however, is
isms in vitro. rather ambiguous by acting either as a tumor promoter
The efficacy of combining sorafenib and doxorubicin (25) or tumor suppressor (24). Notably, we could show
was shown in a phase II trial, resulting in increased for the first time that SPARC is epigenetically up-regu-
overall and progression free survival in patients with lated in human HCC progression by demethylation.
advanced HCC compared to those who received mono- Further studies will reveal the consequences of SPARC
therapy with doxorubicin (23). Furthermore, a phase III overexpression in hepatocarcinogenesis. In the same line,
randomized study which compares sorafenib together aCGH analysis revealed homozygous loss of the Wnt

www.aacrjournals.org Mol Cancer Ther; 10(5) May 2011 859

Downloaded from mct.aacrjournals.org on November 1, 2021. © 2011 American Association for Cancer
Research.
 Published OnlineFirst March 1, 2011; DOI: 10.1158/1535-7163.MCT-10-0917

van Zijl et al.

signaling component AXIN1 as well as of the cation- Disclosure of Potential Conflicts of Interest
selective channel TRPM3. Although deletion of TRPM3
has not been reported in cancer progression, AXIN1 was No potential conflicts of interest were disclosed.
found to be mutated in 5% to 25% of HCC patients and
associated with poor differentiation (26). These findings Acknowledgment
show that our EMT model is genetically well defined and
The authors are grateful to Maria Eisenbauer for excellent technical
thus particularly valid to study human HCC progression assistance.
without unknown genetic variations in widely used
human hepatoma cell lines.
Grant Support
Human models of hepatocellular EMT which reliably
reflect HCC progression are invaluable tools in preclinical This work was supported by the European Union, FP7 Health
studies for (i) the identification of molecular mechanisms Research, project number HEALTH-F4–2008-202047 (W. Mikulits) and
by the Austrian Science Fund, FWF, grant numbers P19598-B13 (W.
underlying HCC progression, (ii) the pharmacological Mikulits), SFB F28 (W. Mikulits) and P20905-B13 (W. Mikulits).
determination of dose–effect relationships and thus the The costs of publication of this article were defrayed in part by the pay-
efficacy of single and combined treatments with novel and ment of page charges. This article must therefore be hereby marked adver-
tisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
currently used anti-cancer drugs, and (iii) the (re)-evalua-
tion of drug target specificity and pleiotropic effects after Received October 1, 2010; revised February 8, 2011; accepted February
drug application. 8, 2011; published OnlineFirst March 1, 2011.

References
1. Thiery JP, Acloque H, Huang RY, Nieto MA. Epithelial-mesenchymal 15. Cheng AL, Kang YK, Chen Z, Tsao CJ, Qin S, Kim JS, et al. Efficacy
transitions in development and disease. Cell 2009;139:871–90. and safety of sorafenib in patients in the Asia-Pacific region with
2. Yang J, Weinberg RA. Epithelial-mesenchymal transition: at the cross- advanced hepatocellular carcinoma: a phase III randomised, double-
roads of development and tumor metastasis. Dev Cell 2008;14:818– blind, placebo-controlled trial. Lancet Oncol 2009;10:25–34.
29. 16. Finn RS. Development of molecularly targeted therapies in hepato-
3. Huber MA, Kraut N, Beug H. Molecular requirements for epithelial- cellular carcinoma: where do we go now? Clin Cancer Res 2010;16:
mesenchymal transition during tumor progression. Curr Opin Cell Biol 390–7.
2005;17:548–58. 17. Llovet JM, Ricci S, Mazzaferro V, Hilgard P, Gane E, Blanc JF, et al.
4. Giannelli G. The epithelial-mesenchymal transition: fact or fiction in Sorafenib in advanced hepatocellular carcinoma. N Engl J Med
cancer? Hepatology 2009;50:1344–6. 2008;359:378–90.
5. van Zijl F, Zulehner G, Petz M, Schneller D, Kornauth C, Hau M, et al. 18. van Zijl F, Mair M, Csiszar A, Schneller D, Zulehner G, Huber H, et al.
Epithelial-mesenchymal transition in hepatocellular carcinoma. Future Hepatic tumor-stroma crosstalk guides epithelial to mesenchymal
Oncol 2009;5:1169–79. transition at the tumor edge. Oncogene 2009;28:4022–33.
6. Yang MH, Chen CL, Chau GY, Chiou SH, Su CW, Chou TY, et al. 19. Weinhaeusel A, Thiele S, Hofner M, Hiort O, Noehammer C. PCR-
Comprehensive analysis of the independent effect of twist and snail in based analysis of differentially methylated regions of GNAS enables
promoting metastasis of hepatocellular carcinoma. Hepatology convenient diagnostic testing of pseudohypoparathyroidism type Ib.
2009;50:1464–74. Clin Chem 2008;54:1537–45.
7. Yuki K, Hirohashi S, Sakamoto M, Kanai T, Shimosato Y. Growth and 20. Sagmeister S, Eisenbauer M, Pirker C, Mohr T, Holzmann K, Zwickl H,
spread of hepatocellular carcinoma. A review of 240 consecutive et al. New cellular tools reveal complex epithelial-mesenchymal inter-
autopsy cases. Cancer 1990;66:2174–9. actions in hepatocarcinogenesis. Br J Cancer 2008;99:151–9.
8. Parkin DM, Bray F, Ferlay J, Pisani P. Global cancer statistics, 2002. 21. Heldin CH, Ostman A, Ronnstrand L. Signal transduction via platelet-
CA Cancer J Clin 2005;55:74–108. derived growth factor receptors. Biochim Biophys Acta 1998;1378:
9. Roxburgh P, Evans TR. Systemic therapy of hepatocellular carci- F79–113.
noma: are we making progress? Adv Ther 2008;25:1089–104. 22. Coulouarn C, Factor VM, Thorgeirsson SS. Transforming growth
10. Bruix J, Llovet JM. Major achievements in hepatocellular carcinoma. factor-beta gene expression signature in mouse hepatocytes predicts
Lancet 2009;373:614–6. clinical outcome in human cancer. Hepatology 2008;47:2059–67.
11. Llovet JM, Bruix J. Novel advancements in the management of 23. Keating GM, Santoro A. Sorafenib: a review of its use in advanced
hepatocellular carcinoma in 2008. J Hepatol 2008;48(Suppl 1):S20– hepatocellular carcinoma. Drugs 2009;69:223–40.
37. 24. Atorrasagasti C, Malvicini M, Aquino JB, Alaniz L, Garcia M, Bolon-
12. Bruix J, Sherman M. Management of hepatocellular carcinoma. trade M, et al. Overexpression of SPARC obliterates the in vivo
Hepatology 2005;42:1208–36. tumorigenicity of human hepatocellular carcinoma cells. Int J Cancer
13. Llovet JM, Bruix J. Systematic review of randomized trials for unre- 2010;126:2726–40.
sectable hepatocellular carcinoma: chemoembolization improves sur- 25. Le Bail B, Faouzi S, Boussarie L, Guirouilh J, Blanc JF, Carles J, et al.
vival. Hepatology 2003;37:429–42. Osteonectin/SPARC is overexpressed in human hepatocellular car-
14. Marelli L, Stigliano R, Triantos C, Senzolo M, Cholongitas E, Davies N, cinoma. J Pathol 1999;189:46–52.
et al. Treatment outcomes for hepatocellular carcinoma using che- 26. Imbeaud S, Ladeiro Y, Zucman-Rossi J. Identification of novel onco-
moembolization in combination with other therapies. Cancer Treat genes and tumor suppressors in hepatocellular carcinoma. Semin
Rev 2006;32:594–606. Liver Dis 2010;30:75–86.

860 Mol Cancer Ther; 10(5) May 2011 Molecular Cancer Therapeutics

Downloaded from mct.aacrjournals.org on November 1, 2021. © 2011 American Association for Cancer
Research.
 Published OnlineFirst March 1, 2011; DOI: 10.1158/1535-7163.MCT-10-0917

A Human Model of Epithelial to Mesenchymal Transition to

Monitor Drug Efficacy in Hepatocellular Carcinoma
Progression
Franziska van Zijl, Sabine Mall, Georg Machat, et al.

Mol Cancer Ther 2011;10:850-860. Published OnlineFirst March 1, 2011.

Updated version Access the most recent version of this article at:
doi:10.1158/1535-7163.MCT-10-0917

Supplementary Access the most recent supplemental material at:

Material http://mct.aacrjournals.org/content/suppl/2011/03/02/1535-7163.MCT-10-0917.DC1

Cited articles This article cites 26 articles, 2 of which you can access for free at:
http://mct.aacrjournals.org/content/10/5/850.full#ref-list-1

Citing articles This article has been cited by 5 HighWire-hosted articles. Access the articles at:
http://mct.aacrjournals.org/content/10/5/850.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications
Subscriptions Department at pubs@aacr.org.

Permissions To request permission to re-use all or part of this article, use this link
http://mct.aacrjournals.org/content/10/5/850.
Click on "Request Permissions" which will take you to the Copyright Clearance Center's
(CCC)
Rightslink site.

Downloaded from mct.aacrjournals.org on November 1, 2021. © 2011 American Association for Cancer
Research.