Loss of runt-related transcription factor 3 expression leads hepatocellular carcinoma cells to escape apoptosis.

Abstract

Runt-related transcription factor 3 (RUNX3) is known as a tumor suppressor gene for gastric cancer and other cancers, this gene may be involved in the development of hepatocellular carcinoma (HCC).
RUNX3 expression was analyzed by immunoblot and immunohistochemistry in HCC cells and tissues, respectively. Hep3B cells, lacking endogenous RUNX3, were introduced with RUNX3 constructs. Cell proliferation was measured using the MTT assay and apoptosis was evaluated using DAPI staining. Apoptosis signaling was assessed by immunoblot analysis.
RUNX3 protein expression was frequently inactivated in the HCC cell lines (91%) and tissues (90%). RUNX3 expression inhibited 90±8% of cell growth at 72 h in serum starved Hep3B cells. Forty-eight hour serum starvation-induced apoptosis and the percentage of apoptotic cells reached 31±4% and 4±1% in RUNX3-expressing Hep3B and control cells, respectively. Apoptotic activity was increased by Bim expression and caspase-3 and caspase-9 activation.
RUNX3 expression enhanced serum starvation-induced apoptosis in HCC cell lines. RUNX3 is deleted or weakly expressed in HCC, which leads to tumorigenesis by escaping apoptosis.

Full text

Page 1

RESEARCH ARTICLEOpen Access
Loss of runt-related transcription factor 3
expression leads hepatocellular carcinoma cells to
escape apoptosis
Yutaka Nakanishi1, Hidenori Shiraha1*, Shin-ichi Nishina1, Shigetomi Tanaka1, Minoru Matsubara1,
Shigeru Horiguchi1, Masaya Iwamuro1, Nobuyuki Takaoka1, Masayuki Uemura1, Kenji Kuwaki, Hiroaki Hagihara,
Junichi Toshimori, Hideki Ohnishi1, Akinobu Takaki1, Shinichiro Nakamura1, Yoshiyuki Kobayashi1,
Kazuhiro Nouso1,2, Takahito Yagi3, Kazuhide Yamamoto1
Abstract
Background: Runt-related transcription factor 3 (RUNX3) is known as a tumor suppressor gene for gastric cancer
and other cancers, this gene may be involved in the development of hepatocellular carcinoma (HCC).
Methods: RUNX3 expression was analyzed by immunoblot and immunohistochemistry in HCC cells and tissues,
respectively. Hep3B cells, lacking endogenous RUNX3, were introduced with RUNX3 constructs. Cell proliferation
was measured using the MTT assay and apoptosis was evaluated using DAPI staining. Apoptosis signaling was
assessed by immunoblot analysis.
Results: RUNX3 protein expression was frequently inactivated in the HCC cell lines (91%) and tissues (90%). RUNX3
expression inhibited 90 ± 8% of cell growth at 72 h in serum starved Hep3B cells. Forty-eight hour serum
starvation-induced apoptosis and the percentage of apoptotic cells reached 31 ± 4% and 4 ± 1% in RUNX3-
expressing Hep3B and control cells, respectively. Apoptotic activity was increased by Bim expression and caspase-3
and caspase-9 activation.
Conclusion: RUNX3 expression enhanced serum starvation-induced apoptosis in HCC cell lines. RUNX3 is deleted
or weakly expressed in HCC, which leads to tumorigenesis by escaping apoptosis.
Background
Hepatocellular carcinoma (HCC)1is the sixth most
common cancer and responsible for more than half a
million deaths worldwide each year [1-3]. Although
most HCC cases occur in East Asia and Middle and
West Africa, its incidence in some developed countries
is increasing [1,4]. In most cases, HCC is fatal because
of an incomplete understanding of the pathogenic
mechanisms and inadequacies of early detection [1,5].
The activation of proto-oncogenes plays a major role
in the development of HCC [1,6-8], and a number of
tumor suppressor genes may be associated with the
development and progression of HCC [1,9-12].
Although several cancer-related genes are altered in
HCC, the frequency of alterations for each individual
gene is relatively low. In HCC, the alteration of tumor
suppressor genes seems to be more important than that
of oncogenes. Established genetic events include the loss
of an allele, mutation, or promoter methylation [13-16].
A higher loss of heterozygosity (LOH) frequency was
detected at several loci on chromosomes 8p23, 4q22-24,
4q35, 17p13, 16q23-24, 6q27, 1p36, and 9p12-14, sug-
gesting the presence of important tumor suppressor
genes at these loci [17]. However, there is little under-
standing of the several key pathways and the genes
involved in these pathways.
Runt-related transcription factor 3 (RUNX3), located
on chromosome 1p36, is correlated with tumorigenesis
and gastric cancer progression [18,19]. RUNX3 acts as
* Correspondence: hshiraha@md.okayama-u.ac.jp
1Department of Gastroenterology and Hepatology, Okayama University
Graduate School of Medicine and Dentistry, 2-5-1 Shikata-cho, Okayama 700-
8558, Japan
Full list of author information is available at the end of the article
Nakanishi et al. BMC Cancer 2011, 11:3
http://www.biomedcentral.com/1471-2407/11/3
© 2011 Nakanishi 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/2.0), which permits unrestricted use, distribution, and
reproduction in any medium, provided the original work is properly cited.

Page 2

an apoptotic factor, downstream of transforming growth
factor-b (TGF-b), and as a cell differentiation mediator
in intestinal metaplasia of gastric mucosa [19-21]. In
gastric cancer cell lines, RUNX3-induced apoptosis
depends on Bim expression [22]. RUNX3 protein
expression is decreased about 45-60% in human gastric
cancer [21] and has been detected in some human
malignancies such as those of the colon, lung, pancreas,
and bile duct [23-26]. RUNX3 gene expression
decreased in 30-80% of HCCs due to LOH and methyla-
tion of its promoter [27,28]. The loss or decrease of
RUNX3 expression in HCC tissue has been recently
reported [29], but the precise function of RUNX3 in
HCC needs to be elucidated.
Methods
Cell lines and cell culture
The HCC cell lines HepG2, Hep3B, PLC/PRF/5 (PLC),
and SK-Hep1 were obtained from the American Type
Culture Collection (Manassas, VA), and the Huh1,
Huh7, JHH1, JHH2, JHH4, HLE, and HLF cell lines
were obtained from the Health Science Research
Resources Bank (Osaka, Japan). Normal human hepato-
cytes were obtained from Sanko Junyaku Co. Ltd.
(Tokyo, Japan). JHH2 and normal human hepatocytes
were cultured in William’s medium E (Invitrogen, Carls-
bad, CA). Other cell lines were maintained in Dulbec-
co’s modified Eagle’s medium (Invitrogen). Media were
supplemented with 10% heat-inactivated fetal bovine
serum (FBS) (Sigma, St. Louis, MO), 1% nonessential
amino acids (Sigma), 1% sodium pyruvate (Sigma), and
1% penicillin/streptomycin solution (Sigma). Cells were
cultured at 37°C in a humidified atmosphere of 5% CO2
and 95% air. Quiescence was carried out under
restricted serum conditions with 0.1% dialyzed FBS for
the indicated time periods.
RNA preparation and reverse transcriptase-polymerase
chain reaction
Total RNA was isolated from cells using Trizol™ reagent
(Invitrogen). Reverse transcription was performed using
random primers and ReverTra Ace™ (Toyobo, Osaka,
Japan) reverse transcriptase (RT). Ps-CA and Ps-CB, pre-
viously published primer set for RUNX3, were utilized
[21]. For each polymerase chain reaction (PCR), 20 μl
(total volume) of reaction mixture contained 0.1 μg tem-
plate DNA, 4 pmol each of the forward and reverse pri-
mers, 2 μl deoxynucleoside triphosphates (200 mM each),
1 U pfu Turbo™ DNA polymerase (Stratagene, La Jolla,
CA), and 2 μl of 10× pfu reaction buffer. PCR amplifica-
tion was conducted on an iCycler™ (Bio-Rad, Hercules,
CA) with the following cycle conditions: cycle 1, 95°C for
2 min; cycles 2-30, 95°C for 30 s, 58°C for 30 s, and 72°C
for 120 s, with a final elongation step of 72°C for 10 min.
Immunoblot analysis
Cells were plated onto 6-well tissue culture plastic dishes
and grown to confluence. After cultivating the cells
under the indicated conditions, they were washed twice
with cold phosphate-buffered saline (PBS) and lysed in
150 μl of sample buffer (100 mM Tris-HCl, pH 6.8, 10%
glycerol, 4% sodium dodecyl sulfate [SDS], 1% bromo-
phenol blue, 10% b-mercaptoethanol). The samples were
resolved by SDS-polyacrylamide gel electrophoresis
(PAGE) and transferred to Immobilon-P™ polyvinylidene
difluoride membranes (Millipore Corporation, Bedford,
MA), which were blocked using Tris-buffered saline with
Tween-20 (TBS-T) (Sigma) containing 5% bovine serum
albumin for 1 h. The membranes were incubated with
antibodies against RUNX3 (R3-G54; Abcam, Cambridge,
MA), poly-histidine (His) (Roche Diagnostics, Basel,
Switzerland), Bax, Bcl-2, Bim, cleaved caspase-3 and -9
(Cell Signaling Technology, Beverley, MA), and b-actin
(Sigma) overnight at 4°C. We washed the membranes
three times with TBS-T and probed with horseradish
peroxidase-conjugated secondary antibodies before devel-
oping them using an ECL Western blotting detection sys-
tem (Amersham Biosciences, Piscataway, NJ) by
enhanced chemiluminescence.
HCC tissue and immunohistochemistry
Thirty-one patients including 24 men with age ranging
from 18 to 71 years (average age, 58 years) and 7 women
with age ranging from 59 to 67 years (average age, 63
years) at the time of hepatic resection were included in
this study. HCC tissues along with adjacent liver tissues
were used for analysis. As per the institutional guidelines,
we obtained informed consent from all donors of liver
tissue samples, and the study was approved by the
Research Ethics Committee of Okayama University.
Immunohistochemistry was performed on formalin-
fixed paraffin sections that were dewaxed and dehy-
drated. After rehydration, endogenous peroxidase activ-
ity was blocked for 30 min in a methanol solution
containing 0.3% hydrogen peroxide. After antigen retrie-
val in citrate buffer, the sections were blocked overnight
at 4°C. The sections were probed with rabbit polyclonal
antibody (ab49117; Abcam) followed by biotinylated
anti-rabbit secondary antibody (Dako Japan, Tokyo,
Japan). The signal was amplified by avidin-biotin com-
plex formation and developed with diaminobenzidine
followed by counterstaining with hematoxylin, after
which the sections were dehydrated in alcohol and
xylene, and mounted for observation. The sections were
scored on a four-tier scale; 0, negative; 1, weak signal; 2,
intermediate signal; and 3, strong signal [30]. All sec-
tions were scored independently by two observers (Y. K.
and K. N.) without prior knowledge. All discrepancies in
scoring were reviewed and a consensus was reached.
Nakanishi et al. BMC Cancer 2011, 11:3
http://www.biomedcentral.com/1471-2407/11/3
Page 2 of 10

Page 3

RUNX3 cloning and transfection
We obtained human RUNX3 cDNA by PCR-based clon-
ing from normal human hepatocytes (Sanko Junyaku).
Briefly, cDNA was amplified by PCR using sense (5’-
TATGCGTATTCCCGTAGA)
CTCGAGGCGGCCGCTCAATGGTGATGGTGAT-
GATGACCGGTACGGTAGGGCCGCCACAC; includ-
ing the six-His tag) oligonucleotide primers with Pfu
Turbo™ Hotstart DNA polymerase (Stratagene) and
cloned into the PCR II TA cloning vector (Invitrogen).
The size of the PCR product was ~1.2 kb. After confir-
mation by sequencing, RUNX3 cDNA was subcloned
into pCEP4 (Stratagene), downstream from a cytomega-
lovirus promoter. The poly-His tag was replaced with
green fluorescent protein (GFP) cDNA from pEGFP-C1
(Clontech, Palo Alto, CA). The human RUNX3 and/or
chloramphenicol acetyltransferase (CAT) (control) con-
structs were transfected into Hep3B cells using
FuGENE™6 transfection reagent (Roche), as per the
manufacturer’s instruction. Cells were selected in com-
plete medium containing 250 μg/ml of hygromycin
(Roche). Polyclonal lines consisting of more than 20
colonies were established. At least two independent sta-
bly transfected lines were established for each construct.
Transient RUNX3 expression was also conducted
using FuGENE™6 in Hep3B, Huh7, HLE, and HLF cells.
After transfection, the cells were cultured under serum
starved condition for the indicated periods, if needed,
and utilized for the following experiments.
andantisense (5’-
MTT assay
Cell proliferative activity was assessed with the 3-(4, 5-
dimethylthiazol-2-yl)-2, 5-diphenyl tetrazolium bromide
(MTT) assay. Briefly, cells were seeded at 2,000 cells/
well in 96-well tissue culture plastic dishes and quiesced
for 6 h with 0.1% dialyzed FBS. After 24-120 h of quies-
cence, the cells were cultured for the indicated periods
with or without 10% FBS. At the end of the treatment,
10 μl of MTT (5 mg/ml in PBS) was added to each well,
and the wells were incubated for an additional 2 h at
37°C. The purple-blue MTT formazan precipitate was
dissolved in 200 μl of dimethyl sulfoxide (Sigma). The
activity of the mitochondria, reflecting cellular growth
and viability, was evaluated by measuring the optical
density at 570 nm with a microplate reader (Bio-Rad).
DAPI staining
Cells were plated at 50% confluence on glass chamber
slides (Labtek II, Nalgen Nunc, Roskide, Denmark) and
quiesced for 6 h with a media containing 0.1% dialyzed
FBS. Then, they were treated with 10% FBS, 100 μM
caspase inhibitor (caspase inhibitor IV, Calbiochem,
Gibbstown, NJ), 1 nM transforming growth factor-a
(TGF-a) (Peprotech Inc. Rocky Hill, NJ), 1 nM
epidermal growth factor (EGF) (Peprotech), and/or 5
ng/ml platelet derived growth factor (PDGF)-BB (Pepro-
tech). Chromosomal DNA was stained with 4’, 6-diami-
dine-2’-phenylindole dihydrochloride (DAPI) (Dojindo,
Kumamoto, Japan) according to the manufacturer’s
instructions. Briefly, treated cells were washed with PBS
and stained with DAPI working solution (1 μg/ml in
PBS) for 2 min. The percentage of cells with condensed
chromatin and/or fragmented nuclei was established in
300-500 DAPI-stained cells examined under a fluores-
cence microscope (IX-70, Olympus, Tokyo, Japan).
Flow cytometry analysis
Annexin V and propidium iodide (PI) staining was per-
formed using an annexin V-fluorescein isothiocyanate
(FITC) Apoptosis Detection kit (Medical & Biological
Laboratories Co., Ltd., Nagoya, Japan) to measure apop-
tosis. Cells were cultured in 10-cm tissue culture plates
and quiesced for 6 h with a media containing 0.1% dia-
lyzed FBS. Cells were cultured in medium with or with-
out 10% FBS for 24 h. Then, they were washed twice
with PBS, collected, and re-suspended in 85 μl of 1×
annexin V-FITC binding buffer. Five microliters of
annexin V-FITC conjugate and 10 ml of PI buffer were
added, and the cells were incubated at room tempera-
ture for 15 min in the dark. After adding 400 μl of 1×
annexin V-FITC binding buffer, cells were analyzed
using a flow cytometer (FACS Calibur; Becton Dickin-
son, Franklin Lakes, NJ).
Gene silencing of Bim with small interfering RNA
RUNX3-expressing Hep3B cells were transfected with
either scrambled negative control small interfering RNA
(siRNA) or Bim siRNA (Applied Biosystems, Foster City,
CA). siRNAs were transfected into cells using RNAi-
Fect™ transfection reagent (Qiagen, Hilden, Germany).
Cells were incubated with scrambled negative control
siRNA or Bim siRNA for 24 h before 48 h of serum
starvation. The MTT assay and DAPI staining for
detecting apoptosis were performed as described above.
Results
Loss of RUNX3 expression in HCC cell lines and human
HCC tissues
A decreased level or absence of RUNX3 mRNA expres-
sion was observed in 10 of 11 HCC cell lines (Figure
1A). RUNX3 mRNA was undetectable in eight cell lines
(HepG2, Hep3B, Huh1, Huh7, JHH1, JHH2, JHH4, and
HLE). In HLF and SK-Hep1 cells, RUNX3 mRNA was
significantly underexpressed (Figure 1A). Normal human
hepatocytes expressed RUNX3 mRNA. Sequence analy-
sis was performed in HLF, PLC, and SK-Hep1 cells, and
no mutation was detected. In accordance with the
mRNA analysis, RUNX3 protein expression was
Nakanishi et al. BMC Cancer 2011, 11:3
http://www.biomedcentral.com/1471-2407/11/3
Page 3 of 10

Page 4

undetectable in the HepG2, Hep3B, JHH1, JHH2, JHH4,
HLE, and HLF cell lines, while the RUNX3 protein was
expressed in HLF, PLC, and SK-Hep1 cells (Figure 1B).
The RUNX3 protein was significantly underexpressed in
HLF and SK-Hep1 cells.
RUNX3 protein expression in human HCC tissue was
compared to that in the corresponding tumor-free
resection margins using immunohistochemical analysis
(Figure 2). Twenty eight (~90%) of these pairs showed a
negative or weak signal for RUNX3 expression in HCC
tissue, but showed RUNX3 protein expression in tumor-
free resection margins (Table 1). In the remaining three
pairs, a weak RUNX3 expression signal was detected in
the tumor-free resection margins; thus, no negative
RUNX3 signal was detected in the tumor-free resection
margins.
Ectopic RUNX3 protein expression in Hep3B cells
To assess whether RUNX3 protein expression affected
cell survival in the HCC cell lines, a RUNX3 construct
was introduced into RUNX3-negative Hep3B cells (Fig-
ure 3A). Overall, the clones were expressed at similar
levels in all cells, as determined by immunocytochemical
analysis (data not shown). RUNX3-expressing Hep3B
cells grew slightly slower than normal Hep3B cells in
the presence of FBS.
RUNX3 expression inhibited cell growth under serum
starvation
RUNX3 has been reported to induce apoptosis in a gas-
tric cancer cell study [21]. The MTT assay was per-
formed to determine whether RUNX3 expression
influenced cell growth. RUNX3-expressing Hep3B cells
grew slightly slower than CAT-transfected Hep3B cells
in the presence of FBS, whereas the growth of RUNX3-
expressing Hep3B cells was markedly suppressed in the
absence of FBS; growth inhibition could be observed as
early as 24 h, and reached 70 ± 12% and 90 ± 8% at 48
and 72 h, respectively (Figure 3B). The inhibition levels
were over 4 times than those found in the condition
with 10% FBS. This effect was confirmed with GFP-
tagged RUNX3-expressing Hep3B cells (70 ± 11%
growth inhibition at 72 h).
RUNX3 expression induced apoptosis under serum
starvation
The effect of RUNX3 expression on cell survival and the
cell cycle with and without FBS was assessed to investi-
gate whether the elicited growth suppression in
RUNX3-expressing cells under serum starved conditions
was due to an increase in cell death or due to cell cycle
inhibition, or both. DAPI staining demonstrated that
serum starvation induced apoptosis in RUNX3-expres-
sing Hep3B cells (31 ± 4%) but not in CAT-transfected
Hep3B cells (4 ± 1%) in the absence of FBS (Figure 3C).
Flow cytometry analysis with annexin V antibody was
also performed. RUNX3-expressing Hep3B cells showed
a significant increase in a pre-apoptosis population
(Annexin V+ PI-) after 24 h of serum starvation com-
pared with CAT-transfected Hep3B cells (Figure 3D).
RUNX3-induced apoptosis through the Bim-caspase
pathway
Because a RUNX3-induced apoptotic pathway has been
described previously, the effect of altering RUNX3
expression was investigated. Bim protein expression was
Figure 1 RUNX3 mRNA (A) and protein (B) expression in HCC
cell lines. (A) RUNX3 and GAPDH mRNA expression levels were
determined by RT-PCR. Shown here are representative gels from
three independent experiments. (B) RUNX3 and a-actin protein
expression was analyzed by immunoblot using the anti-RUNX3
antibody. Shown here are representative blots from more than
three independent experiments.
Figure 2 Immunohistochemical staining of RUNX3 in human
liver tissues. (A) nontumor liver tissue with a protein score of 3 (B)
HCC tissue with a protein score of 0. Bar = 100 μm.
Table 1 RUNX3 expression in HCC samples (n = 31) and
the corresponding tumor-free resection margins
RUNX3 protein
expression score= 31)
HCC samples (nTumor-free sections
(n = 31)
0 (negative signal)
1 (weak signal)
2 (intermediate signal)
3 (strong signal)
13 (41.9%)
15 (48.4%)
3 (9.7%)
0
0
3 (9.7%)
16 (51.6%)
12 (38.7%)
Nakanishi et al. BMC Cancer 2011, 11:3
http://www.biomedcentral.com/1471-2407/11/3
Page 4 of 10

Page 5

enhanced by serum starvation in RUNX3-expressing
Hep3B cells but not in control cells (Figure 4A). Acti-
vated apoptosis executors, caspase-9 and -3, were found
in serum starved RUNX3-expressing Hep3B cells.
Expression of the Bim attenuators, Bax and Bcl-2, was
not affected by serum starvation. These results imply
that Bim plays a major role in serum starvation-induced
apoptosis in RUNX3-expressing cells.
Serum starvation-induced apoptosis was abrogated by
an apoptosis inhibitor (Figure 4B). Various growth fac-
tors were employed to determine whether serum star-
vation-induced apoptosis was caused by the absence of
a growth factor-induced survival signal. As a result,
TGF-a, EGF, and PDGF abrogated serum starvation-
induced apoptosis in RUNX3-expressing Hep3B cells
(Figure 4B).
Figure 3 RUNX3 expression in Hep3B cells. CAT (control) and RUNX3 constructs were introduced into Hep3B cells. Polyclonal stably
expressing cell lines were selected in the presence of hygromycin. (A) Immunoblot analysis RUNX3 protein expression was analyzed by
immunoblot using anti-RUNX3 and anti-His antibodies. Shown here are representative blots from more than three independent experiments. (B)
Effect of RUNX3 expression on cell growth Cell proliferative activities were measured by the MTT assay. All results are expressed as ratios to cell
number at day 0. Data represent the mean ± S.E. of more than three independent experiments, each with four replicates. *, P < 0.05; **, P < 0.01
(vs. data at 0 h); Student’s t-test. (C) Apoptosis determined by DAPI staining Cells were cultured in medium without FBS for the indicated
periods. The cells were stained with DAPI, and the percentage of apoptotic cells was examined by fluorescence microscopy (CAT; white bars,
RUNX3; black bars). Data represent the mean ± S.E. of more than five independent experiments, each with triplicates. n.s., not significant; **, P <
0.01 (vs. data at 0 h); Student’s t-test. (D) Flow cytometry analysis Cells were cultured in medium with or without 10% FBS for 24 h. They were
harvested and washed in PBS without Ca2+/Mg2+, and stained using the Mebcyto™ apoptosis detection kit. The cells were analyzed with a
Becton Dickinson FACS Calibur flow cytometer. Shown here are representative plots from more than three independent experiments.
Nakanishi et al. BMC Cancer 2011, 11:3
http://www.biomedcentral.com/1471-2407/11/3
Page 5 of 10