Protein Phosphatase 2A promotes Hepatocellular Carcinogenesis in the Diethylnitrosamine Mouse Model through Inhibition of p53.

Abstract

Hepatocellular carcinoma (HCC) is one of the most common cancers worldwide. Most HCCs develop in cirrhotic livers. Alcoholic liver disease, chronic hepatitis B and chronic hepatitis C (CHC) are the most common underlying liver diseases. HCV-specific mechanisms that contribute to HCC are presently unknown. Transgenic expression of HCV proteins in the mouse liver induces an over-expression of the protein phosphatase 2A catalytic subunit (PP2Ac). We have previously reported that HCV induced PP2Ac over-expression modulates histone methylation and acetylation and inhibits DNA damage repair. In the present work, we analyze tumor formation and gene expression using HCV transgenic mice that over-express PP2Ac and liver tissues from HCC patients. We demonstrate that PP2Ac over-expression interferes with p53-induced apoptosis. Injection of the carcinogen diethylnitrosamine induced significantly more and larger liver tumors in HCV transgenic mice that over-express PP2Ac compared to control mice. In human liver biopsies from patients with HCC, PP2Ac expression was significantly higher in HCC tissue compared to non-tumorous liver tissue from the same patient. Our findings demonstrate an important role of PP2Ac over-expression in liver carcinogenesis and provide insights into the molecular pathogenesis of HCV-induced HCC.

Full text

© The Author 2013. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com
Carcinogenesis vol.35 no.1 pp.114–122, 2014
doi:10.1093/carcin/bgt258
Advance Access publication August 5, 2013
Protein phosphatase 2A promotes hepatocellular carcinogenesis in the
diethylnitrosamine mouse model through inhibition of p53
François H.T.Duong1, Michael T.Dill1,2, Matthias
S.Matter3, ZuzannaMakowska1, DiegoCalabrese1,
TanjaDietsche3, SylviaKetterer1, LuigiTerracciano3 and
Markus H.Heim1,2,*
1Department of Biomedicine and 2Division of Gastroenterology and
Hepatology, University Hospital Basel, CH-4031 Basel, Switzerland and
3Department of Molecular Pathology, Institute for Pathology, University
Hospital Basel, CH-4003 Basel, Switzerland
*To whom correspondence should be addressed. Division of Gastroenterology
and Hepatology, University Hospital Basel, Petersgraben 4, CH-4031 Basel,
Switzerland. Tel:+41 61 265 33 62; Fax:+41 61 265 53 52;
Email: markus.heim@unibas.ch
Hepatocellular carcinoma (HCC) is one of the most common can-
cers worldwide. Most HCCs develop in cirrhotic livers. Alcoholic
liver disease, chronic hepatitis B and chronic hepatitis C are
the most common underlying liver diseases. Hepatitis C virus
(HCV)-specic mechanisms that contribute to HCC are presently
unknown. Transgenic expression of HCV proteins in the mouse
liver induces an overexpression of the protein phosphatase 2A cat-
alytic subunit (PP2Ac). We have previously reported that HCV-
induced PP2Ac overexpression modulates histone methylation
and acetylation and inhibits DNA damage repair. In this study, we
analyze tumor formation and gene expression using HCV trans-
genic mice that overexpress PP2Ac and liver tissues from patients
with HCC. We demonstrate that PP2Ac overexpression interferes
with p53-induced apoptosis. Injection of the carcinogen, diethyl-
nitrosamine, induced signicantly more and larger liver tumors
in HCV transgenic mice that overexpress PP2Ac compared with
control mice. In human liver biopsies from patients with HCC,
PP2Ac expression was signicantly higher in HCC tissue com-
pared with non-tumorous liver tissue from the same patients. Our
ndings demonstrate an important role of PP2Ac overexpression
in liver carcinogenesis and provide insights into the molecular
pathogenesis of HCV-induced HCC.
Introduction
Hepatocellular carcinoma (HCC) is the most prevalent primary liver
cancer and a major cause of cancer mortality worldwide. The major
risk factor for HCC is chronic hepatitis B, followed by chronic hepa-
titis C (CHC) and chronic alcoholic liver disease (1). In Europe, most
HCCs occur in cirrhosis, and it is well established that cirrhosis is a
major risk factor for HCC independent of the underlying liver dis-
ease. However, in the context of chronic hepatitis B or CHC, HCC can
develop also in non-cirrhotic livers. Virus-induced chronic inamma-
tion can impair DNA damaged repair, rendering cells more suscepti-
ble to spontaneous or mutagen-induced alterations (2). There are also
experimental evidences of the oncogenic potential of hepatitis C virus
(HCV) proteins such as HCV core. For instance, it has been reported
that HCV core transgenic mice develop hepatic steatosis and HCC at
the age of 16–19months (3).
We have reported previously that HCV infection induces an upreg-
ulation of protein phosphatase 2A (PP2A) catalytic subunit through
an endoplasmic reticulum stress response pathway (4,5). PP2A is a
heterotrimeric protein phosphatase consisting of a 36 kD catalytic
C subunit (PP2Ac), a 65 kD structural A subunit and a variable
regulatory B subunit. This phosphatase is widely expressed in all cell
types and is involved in the posttranslational control of signaling pro-
teins by performing reversible phosphorylation (6). In recent work,
we observed that HCV-induced PP2Ac overexpression in cells leads
to an inhibition of histone H4 posttranslational modications and
impairs the DNA repair machinery (7). We also found that cotrans-
fection of c-myc and HA-PP2Ac plasmids into NIH3T3 cells sig-
nicantly increased c-myc-induced anchorage-independent growth.
These data suggested that PP2Ac overexpression may contribute to
hepatocellular carcinogenesis in the context ofCHC.
Cells are constantly exposed to DNA damage. In order to pre-
vent the propagation of mutations, cells have evolved systems that
sense DNA damage and interrupt the cell cycle to allow repair of
damaged DNA. If DNA damage is too severe and irreparable, cells
undergo apoptosis (8). DNA damage is a hallmark of cancer cells.
DNA damage repair is often ineffective and proapoptotic pathways
are non-functional in many cancer types, resulting in an accumula-
tion of mutations during cancer evolution (9). PP2A is of particular
interest in the control of apoptotic networks because of its predomi-
nant tumor-suppressive characteristics (10). However, some reports
have shown that particular PP2A complexes can counterbalance the
apoptotic cell death program as well. Indeed, it has been described
that cyclin G associates with PP2A holoenzyme containing B’ subu-
nits and directs the phosphatase complex to the murine double minute
2 (Mdm2), where PP2A dephosphorylates Mdm2 on threonine 216,
leading to a negative regulation of p53 apoptotic function (11).
The tumor suppressor protein p53 can initiate cell cycle arrest, DNA
repair and apoptosis (12,13). In resting cells, constant degradation
through the proteasome pathway maintains low expression levels of
p53. Upon DNA damage, p53 is stabilized and accumulated (14). This
stabilization results from phosphorylation of p53 on serine 15 within
the transactivation domain, inhibiting the association to Mdm2, thereby
preventing p53 ubiquitination and proteasomal degradation (15).
Several functions of p53 are also regulated by phosphorylation (16,17).
For instance, phosphorylation of p53 on serine 37 and 46 is associated
with induction of transcriptional activity and apoptosis (18,19). p53 is
frequently found to be mutated in human cancer (20,21), and malig-
nancies that retain a wild-type p53 gene usually acquire other impair-
ments that affect p53 function (22), suggesting that p53 inactivation is
a common step in most human cancers development.
In this study, we investigated p53 modications and apoptotic
pathways in PP2Ac-overexpressing cells and HCVcc-infected cells.
We show that PP2Ac overexpression reduces p53 phosphorylation in
response to DNA-damaging agents. Consequently, expression of p53
proapoptotic genes is reduced, and apoptosis is inhibited. HCV trans-
genic mice that overexpress PP2Ac were more susceptible to diethyl-
nitrosamine (DEN) and had more and larger liver tumors. In human
liver biopsies from patients with CHC and HCC, we found an overex-
pression of PP2Ac in tumor tissue and observed a signicant negative
correlation between PP2Ac and PUMA/Noxa expression.
Materials and methods
Cells and reagents
UHCV57.3, HA-PP2Ac and short-hairpin small interfering RNA PP2Aca
cells were described previously (4,7,23). Etoposide and DEN were purchased
from Sigma–Aldrich (Fluka Chemie GmbH, Buchs, Switzerland). Etopophos
was obtained from Bristol-Myers Squibb SA (Baar, Switzerland). Anti-
HCV core (clone C7-50) was from ABR Afnity Bioreagents (Lucerna
Chem AG, Lucerne, Switzerland). Anti-Caspase-9, cleaved Caspase-3,
pSer15p53, pSer37p53, pSer18p53 and pSer46p53 were from Cell Signaling
(Bioconcept, Allschwil, Switzerland). Anti-PP2Ac was from Millipore AG
(Zug, Switzerland). Anti-p53 was from Santa Cruz Biotechnology (LabForce
AG, Nunningen, Switzerland).
Abbreviations: CHC, chronic hepatitis C; CREB, cyclic adenosine
monophosphate response element-binding protein; DEN, diethylnitrosamine;
HCC, hepatocellular carcinoma; HCV, hepatitis C virus; PP2Ac, protein phos-
phatase 2A catalytic subunit; SEM, standard error of the mean.
114
Downloaded from https://academic.oup.com/carcin/article-abstract/35/1/114/360222
by guest
on 06 November 2017

PP2Ac upregulation increases DEN-mediated liver tumors
HCV particles production and infection of Huh7.5.1cells
RNA preparation and electroporation were done as described (24). Huh7.5.1
cells were infected with HCV particles at multiplicity of infection of 1 for
3days.
Animals and experimentaldesign
HCV transgenic mice (B6HCV) were described previously (25). The DEN
injection study was performed in two groups of mice: C57BL6 injected with
DEN (5 µg/g) (n = 22) and B6HCV injected with DEN (5 µg/g) (n=19). DEN
was injected once intraperitoneally to mice at 2 weeks of age. The animals
were euthanized 40 weeks after injection by CO2 inhalation. The resected
liver lobes were immediately frozen in liquid nitrogen and kept at −70°C. The
Etopophos injection study was performed in four groups of mice at 8 weeks of
age: C57BL6, no injection (n=2); C57BL6 injected with Etopophos (100 mg/
m2) (n=2); B6HCV, no injection (n=2) and B6HCV injected with Etopophos
(100 mg/m2) (n=2). Etopophos was injected intraperitoneally to the mice and
the animals were euthanized after 2 h.
Microcystin LR (Sigma–Aldrich, Fluka Chemie GmbH) injection (70 µg/
kg) was performed intraperitoneally, and the mice were euthanized after 2 h.
Protein preparation and western blotting analysis
Whole cell extracts (from cell lines and from mouse liver tissues) and
western blot analysis were performed as described previously (7,26).
Densitometry analysis of protein bands was performed with ImageJ soft-
ware (NIH Image).
RNA isolation, reverse transcription and SYBR–PCR
RNA isolation, reverse transcription and SYBR–PCR were done as described
previously (7). The human primers were 5′-TCTTCCTCTGAGGCGAGCT-3′
and 5′-AGGTGTGTGTGTCTGAGCCC-3′ for p53AIP1; 5′-AGAGCTGGAA
GTCGAGTGT-3′ and 5′-GCACCTTCACATTCCTCTC-3′ for Noxa; 5′-AAG
ACCATGTGGACCTGT-3′ and 5′-GGTAGAAATCTGTCATGCTG-3′ for p21;
5′-CCTGCCAGATTTGTGAGACAAG-3′ and 5′-CAGGAGTCCCATGATG
AGATTG-3′ for PUMA; 5′-CCACAGCAAGTCACACATTGG-3′ and 5′-CAG
AGCACTTGATCGCCTACAA-3′ for PP2Ac; 5′-GCTCCTCCTGTTCGACA
GTCA-3′ and 5′-ACCTTCCCCATGGTGTCTGA-3′ for GAPDH. The mouse
primers were 5′-ATCCGCAAGCCTGTGACTGT-3′ and 5′-TCGGGCCAGG
GTGTTTTT-3′ for RPL19; 5′-GCCCAGCAGCACTTAGAGTC-3′ and 5′-TGT
CGATGCTGCTCTTCTTG-3′ for PUMA; 5′-CCCAGATTGGGGACCTTA
GT-3′ and 5′-TCCTCATCCTGCTCTTTTGC-3′ for Noxa.
Paired liver biopsies
All patients were recruited in the Hepatology Outpatient Clinic of the
University Hospital Basel, Switzerland. All patients gave written informed
consent to participate in this study and donated a liver biopsy specimen
for research purposes. The study was approved by the Ethics Committee
ofBasel.
Biopsies of the tumor and non-tumorous tissue from a distal site of the liver
were performed using coaxial technique. One biopsy specimen from each site
was assessed by a pathologist. Biopsy specimens with at least 50% of HCC
tissue in the tumor sample and tumor free in the control sample were used for
the analysis.
Biopsy samples were frozen immediately after collection and stored in
liquid nitrogen until processing. Total RNA was extracted using Qiazol
reagent and RNAeasy Mini Kit (Qiagen) according to the manufacturer’s
instructions. Reverse transcription and quantitative PCR were performed as
described above.
Soft agar assay
Human hepatoma cells and human primary lung broblasts (kindly pro-
vided by Prof. Michael Roth) transfected with HA-PP2Ac plasmid using
FuGene HD, according to the manufacturer’s recommendations, were cul-
tured for 7days in a semisolid culture media, and colonies were visualized
under light microscopy.
Caspase-9assay
Huh7 and HA-PP2Ac cells were treated with 100 µM etoposide for
1 h, and caspase-9 activity was measured using Caspase-Glo 9 Assay
(Promega AG, Dübendorf, Switzerland) according to the manufacturer’s
instructions.
In vitro dephosphorylation assay
Whole cell lysate from Huh7 cells stimulated with 100 µM etoposide for
1 h was prepared and then incubated with or without 1 U of puried PP2A
(Upstate, Millipore AG) at 37°C for 10 min. The reaction was then stopped by
heating at 100°C for 5 min and then loaded on a sodium dodecyl sulfate–poly-
acrylamide gel electrophoresis. pSer15p53, pSer37p53 and pSer46p53 were
then visualized with specic antibodies.
Cleaved caspase-3 immunouorescence staining
Briey, sections (10 µm) from mouse liver embedded in optimal cutting
temperature compound were prepared and mounted onto polylysine slides
(Thermo Fisher Scientic, Wohlen, Switzerland). Tissue sections were then
xed in absolute methanol (−20°C, 15 min) and then rinsed twice in phos-
phate-buffered saline before being blocked in a 2% bovine serum albumin
and 0.2% Triton-X solution for 30 min at room temperature. Primary anti-
body incubation was performed overnight at 4°C. Slides were then rinsed
twice in phosphate-buffered saline. Epitopes detection was performed using
uorescein-conjugated goat anti-rabbit (Invitrogen Molecular Probes, Eugene,
OR). Nuclei were stained by 4′,6-diamidino-2-phenylindole, and slides
were mounted with a water-based mounting medium (Ultramount Aqueous
Permanent Mounting Medium; Dako, Carpinteria, CA).
Parafn-embedded sections and hematoxylin staining
Fresh liver tissue was xed in 4% paraformaldehyde overnight. Afterwards,
tissue was embedded in parafn, and the sections of 4 µm thickness were
prepared, stained with H&E and analyzed under light microscopy. Liver
tumors were classied according to a consensus report on murine liver
lesions (27).
Results
PP2Ac overexpression impairs p53 phosphorylation, p53-induced
target genes expression and p53-induced apoptosis
We have shown previously that the expression of HCV proteins in
cells induces PP2Ac protein expression (4). To explore a potential
role of PP2A in the regulation of p53 function, we used cells that
allow the inducible expression of HCV proteins (UHCV57.3 cells)
to upregulate PP2Ac and then stimulated p53 phosphorylation with
the DNA-damaging agent etoposide. As shown in Figure 1, HCV-
mediated upregulation of PP2Ac impaired etoposide-induced p53
phosphorylation on serine 37 and 46 (Figure 1A) and signicantly
reduced expression of the canonical p53 target genes—p53AIP1,
Noxa, p21 and PUMA (Figure1B). Next, we conrmed the role of
PP2Ac in Huh7-derived cells that express a constitutive active form of
PP2Ac (HA-PP2Ac). Compared with control Huh7 cells, etoposide-
induced phosphorylation of serine 37 and 46 was signicantly inhib-
ited (Figure1C). Accordingly, the induction of p53 target genes was
also reduced in HA-PP2Ac (Figure1D).
There was also a slight reduction of serine 15 phosphorylation
in UHCV57.3 cells but not in HA-PP2Ac cells. To assess if HCV
causes an additional inhibition of p53 phosphorylation on serine 15,
we infected Huh7.5 cells with the cell culture infectious JC-1 clone
of HCV. HCV-infected cells indeed had an overexpression of PP2Ac,
and etoposide-induced phosphorylation of serine 37 and 46 was inhib-
ited (Supplementary Figure S1, available at Carcinogenesis Online).
However, no change in serine 15 phosphorylation was observed.
We further explored the mechanism by which PP2A reduces etopo-
side-induced p53 phosphorylation using an in vitro dephosphorylation
assay. Extracts from etoposide-treated Huh7 cells were incubated with
puried PP2A for 10 min, followed by western blotting using antibod-
ies specic for phosphorylated serine residues on p53. There was a
strong reduction of phosphoserine 37 and 46 but no effect on serine 15
(Supplementary Figure S2, available at Carcinogenesis Online).
Activated p53 can eliminate tumor cells by inducing apoptosis
through caspase-9 activation. We, therefore, measured caspase-9
activity following etoposide treatment and found that the activity
of caspase-9 was significantly reduced in HA-PP2Ac cells
compared with Huh7 cells (Figure 2A), suggesting that PP2Ac
overexpression could favor the survival of transformed cells.
Next, we transfected human primary lung fibroblasts with the
HA-PP2Ac plasmid and performed a cell anchorage-independent
growth assay. In line with our data that were published previously
(7), PP2Ac overexpression resulted in transformation of human
primary lung fibroblasts as illustrated by colony formation in soft
agar media (Figure2B).
The role of PP2A in the regulation of p53 function was fur-
ther explored by silencing PP2Ac using short-hairpin small inter-
fering RNA. Knockdown of PP2Ac resulted in an increased
115
Downloaded from https://academic.oup.com/carcin/article-abstract/35/1/114/360222
by guest
on 06 November 2017

F.H.T.Duong etal.
Fig.1. PP2Ac overexpression impairs p53 phosphorylation and expression of p53 proapoptotic genes. (A) UHCV57.3 cells were derepressed by tetracycline
for 24 h in order to induce the expression of HCV proteins. Cells were then exposed to 100 µM etoposide for 1 h, and phosphorylation of p53 on serine 15,
37 and 46 was analyzed by western blotting. (B) Cells were stimulated with 100 µM etoposide for 24 h, and total RNA was isolated and reverse transcribed to
complementary DNA. The expression of genes was measured by quantitative PCR (qPCR). Results are expressed as mean ± standard error of the mean (SEM)
from three independent experiments. Statistical analysis was performed using Student’s t-test. (C) Huh7 and HA-PP2Ac cells were exposed to 100 µM etoposide
for 1 h, and phosphorylation of p53 on serine 15, 37 and 46 was analyzed by western blotting. (D) Huh7 and HA-PP2Ac cells were exposed to 100 µM etoposide
for 24 h, and RNA was isolated. The expression of genes was monitored by qPCR. Results are expressed as mean ± SEM from three independent experiments.
Fig.2. PP2Ac overexpression reduces caspase-9 activation upon etoposide stimulation and promotes broblast transformation. (A) Huh7 and HA-PP2Ac cells
were stimulated with 100 µM etoposide for 1 h, and caspase-9 activity was measured using Caspase-Glo 9 Assay. Results are expressed as mean ± SEM from
two independent experiments. (B) Huh7 cells or human broblasts untransfected or transfected with an HA-PP2Ac coding plasmid were cultured on a semisolid
culture media for 7days. Colonies were visualized by light microscopy. Representative pictures of colonies after 7days of culture on semisolid support are
shown.
116
Downloaded from https://academic.oup.com/carcin/article-abstract/35/1/114/360222
by guest
on 06 November 2017

PP2Ac upregulation increases DEN-mediated liver tumors
etoposide-induced phosphorylation of p53 on serine 37 and 46, but
not on serine 15 (Figure3A), and consequently a signicantly higher
expression of Noxa and PUMA, two direct p53-induced apoptotic
genes (Figure3B).
PP2Ac overexpression in B6HCV mice impairs Etopophos-
mediated p53 phosphorylation, reduces DEN-induced p53 target
gene expression and increases DEN-induced HCC formation
To assess the in vivo relevance of our ndings, we injected B6HCV
and C57BL6 control mice with Etopophos. We used B6HCV mice
because no transgenic mouse model with direct PP2Ac overexpres-
sion exists, but B6HCV mice consistently showed a signicant over-
expression of PP2Ac in the liver (4). Upon Etopophos injection, we
observed phosphorylation of p53 on serine 18 (homologue to serine
15 in human) in liver homogenates of C57BL6 mice. This phospho-
rylation is impaired in B6HCV mice (Figure4A). We were not able
to investigate phosphorylation on serine 46 and 37 because no anti-
bodies against these phosphorylation sites are available for mouse.
We investigated further the effect of PP2A on p53-mediated apopto-
sis by analyzing caspase-9 cleavage upon injection with microcystin
LR, a DNA damage compound. As shown in Supplementary Figure
S3, available at Carcinogenesis Online, a clear reduction of cleaved
caspase-9 in B6HCV mice can be observed, suggesting a reduction of
apoptosis in these transgenic animals.
Carcinogenesis is a complex process involving alterations of DNA
damage repair and apoptosis (28). We have shown previously that
PP2Ac overexpression renders cells more susceptible to DNA damage
and impairs DNA repair (7). In the present paper, we report an
impairment of apoptosis. We, therefore, investigated if B6HCV mice
that overexpress PP2Ac are more susceptible to HCC. B6HCV and
C57BL6 mice were injected with DEN, a carcinogen known to induce
liver tumors in mice. Mice were euthanized 40 weeks after a single
injection of a low dose of DEN. We observed a reduction of Noxa and
PUMA, two direct p53 target genes (Figure5A) in non-tumoral liver
tissue obtained from B6HCV mice, suggesting a diminution of p53
proapoptotic function. We, therefore, analyzed cleaved caspase-3 in
B6HCV and in C57BL6 mice. Our data show a reduction of caspase-3
cleavage in B6HCV mice compared with control animals, suggesting
a diminution of DEN-induced apoptosis in the transgenic mice
(Figure5B). We conrmed these results by staining cleaved caspase-3
on liver sections. As shown in Figure5C, C57BL6 mice showed more
caspase-3 cleavage than B6HCV mice. Furthermore, we observed
pronounced chromatin condensation in control animals compared
Fig.3. PP2Ac silencing alters p53 phosphorylation and p53-induced proapoptotic genes expression. (A) Control-scrambled and short-hairpin small interfering
RNA PP2Ac cells were treated with 100 µM etoposide for 1 h, and phosphorylation on p53 was monitored by western blotting. A representative blot from at least
two independent experiments is shown. (B) Cells were treated with 100 µM etoposide for 24 h, and the expression of the genes was quantied by quantitative
PCR. Results are expressed as mean ± SEM from three independent experiments.
Fig.4. PP2Ac overexpression in HCV transgenic mice impairs p53
phosphorylation upon Etopophos injection. Eight-week-old C57BL6 (n = 2)
and B6HCV (n = 2) mice were injected with Etopophos and killed 2 h later.
Liver homogenates were analyzed by western blotting with antibodies to
phosphoserine 18 on p53, p53, actin, HCV core and PP2Ac.
117
Downloaded from https://academic.oup.com/carcin/article-abstract/35/1/114/360222
by guest
on 06 November 2017

F.H.T.Duong etal.
Fig.5. PP2Ac overexpression in HCV transgenic mice reduces p53 proapoptotic function and enhances liver tumor formation upon DEN injection. (A) C57BL6
and B6HCV mice were injected intraperitoneally with DEN at 5 µg/g of body weight. Animals were killed 40 weeks after DEN administration. The expression
of Noxa and PUMA was quantied by quantitative PCR and expressed relative to RPL19. Results are expressed as mean ± SEM from 22 control animals and 17
transgenic mice, and statistical analysis was performed using Student’s t-test. (B) Cleaved caspase-3 was detected by immunoblotting in liver homogenates from
mice injected with DEN. A representative blot is shown. (C) Detection of caspase-3 cleavage by immunouorescence from mouse liver sections. A representative
118
Downloaded from https://academic.oup.com/carcin/article-abstract/35/1/114/360222
by guest
on 06 November 2017

PP2Ac upregulation increases DEN-mediated liver tumors
with transgenic mice, demonstrating more apoptosis in C57BL6
mice (Figure 5C). We counted more liver tumors in B6HCV mice
than in C57BL6 mice (Figure5D). These liver tumors were bigger
in B6HCV mice than in control mice (Figure5D and 5E). The body
weight and the liver weight remained unchanged between control
and transgenic mice (Figure5D). Histology of the liver tumors did
not show any signicant difference between B6HCV and C57BL6
mice (Figure5E). In both groups, the tumors were characterized by
loss of the normal lobular architecture and displayed irregular growth
pattern. They were sharply demarcated from the surrounding liver
parenchyma that was often compressed. Moreover, focal areas of
cellular atypia characterized by slight pleomorphic nuclei, coarsely
clumped chromatin, large nucleoli and cytoplasmic basophilia were
also present. All these histological features are consistent with murine
hepatocellular adenomas, which are typically found after a single
DEN injection (29).
PP2Ac is upregulated in tumorous liver biopsies and negatively
correlated with p53 proapoptotic gene expression
In order to assess the correlation between PP2Ac expression and p53
proapoptotic gene expression in humans, we performed a quantita-
tive PCR analysis using paired liver biopsies from HCC and from tis-
sue outside the tumors collected from patients with and without CHC
(Table I). Indeed, analysis of PP2Ac expression revealed a higher
expression of PP2Ac in malignant tissue than in non-tumorous sur-
rounding tissue both from non-CHC patients (Figure 6A). We then
Table I. Clinical characteristics of HCC liver biopsies
No. Code Age Sex Fibrosis
grade
HCV/HBV infection Alcohol
abuse
Focality Pathology Edmonson
grade
1 A1 63 M F4 No virus Yes Multifocal Trabecular, pseudoglandular,
inammation
III
2 A12 69 M F0 No virus No Singular Trabecular, focally necrotic II
3 A37 68 M F4 No virus Yes Singular Trabecular-sinusoidal II
4 A255 79 F F3–F4 HBV No Singular Trabecular II
5 A268 63 M F4 No virus Yes Multifocal Trabecular, hyaline globules II
6 A286 63 M F3 No virus – Singular Trabecular, lamellar brosis,
steatosis
II
7 A367 64 M F3–F4 No virus Yes Singular Trabecular, focally clear cell II
8 A442 75 M F4 No virus Yes Multifocal NA II
9 A722 80 F F1 HCV GT NA No Singular Trabecular II
10 A724 60 M F4 HBV No Multifocal Mixed hepatocholangiocarcinoma III
11 A824 62 M F3–F4 HBV Yes Singular Mixed hepatocholangiocarcinoma III
12 A829c 30 M F1–F2 HBV No Singular Stem cell-like, mixed
hepatocholangiocarcinoma
IV
13 A835 56 M F4 HCV GT 3a Yes Three
nodules
Trabecular, steatosis,
desmoplastic
III
14 A879 67 M F3–F4 HBV – Multifocal Trabecular II
15 A896 71 M F4 No virus Ye s Multifocal Trabecular III
16 A911 52 M F4 HCV GT NA – NA Pseudoglandular II
17 A915 85 M F0–F1 No virus No Singular Trabecular, pseudoglandular II
18 A941 48 M F4 HCV GT 1a (6.25 log10) Yes Multifocal Trabecular, partly necrotic III
19 A969 68 M F4 No virus No Singular Trabecular II
20 B21 75 M F4 No virus Yes Singular Trabecular, Mallory bodies,
granulomatous inammation
II–III
21 B39 59 M F4 No virus Yes Singular Trabecular, inammation III
22 B62 58 F F4 HBV – Singular Trabecular II (focally I)
23 B77 59 M F4 No virus Yes Multifocal Solid III
24 B156 78 M F4 No virus Yes Multifocal Trabecular, microacinary II
25 B214a 75 M F3 No virus Ye s Two foci Solid II
26 B277 62 M F4 HCV GT 1b (4.79 log10) – Singular Solid, necrosis, inammation III
27 B313 69 F F4 HBV No Two foci Trabecular, inammation II
28 B322 65 M F4 No virus Yes Two foci Trabecular, inammation II
29 B326 54 M F4 HCV GT 1b (5.18 log10) Yes Singular Trabecular III
30 B370 60 M F0 No virus Yes Multifocal Trabecular, focally clear cell II
31 B399 52 M F4 HCV GT 1b (4.23 log10) Yes Multifocal Trabecular III
32 B410 76 M F4 No virus Yes Multifocal Trabecular II
33 B442 57 M F4 No virus Yes Multifocal Trabecular, lamellary brosis III
34 B459 50 M F4 HBV No Singular Trabecular II
35 B540 63 M F4 No virus Yes Multifocal Trabecular, inammation,
steatosis
II
36 B568 85 M NA No virus Ye s Three foci Trabecular, partly necrotic II
37 B592 76 M F4 No virus Yes Multifocal,
diffuse
Trabecular, hyaline globules II
38 B597 57 M F4 No virus Yes Singular Trabecular II
HBV, hepatitis B virus.
result is shown. Right panel shows enlargement of delimitated areas. Chromatin condensation, an apoptotic nuclear feature, is clearly visible in C57BL6 mouse
that received DEN. (D) Body weight, liver weight, number and size of macroscopic lesions recognizable from the outside were measured 40 weeks after DEN
injection. Results are expressed as mean ± SEM from 22 control mice and 17 HCV transgenic animals. (E) Macroscopic and microscopic examples of liver
tumors from C57BL6 and B6HCV mice 40 weeks after receiving 5 µg/g of DEN. Liver tumors were easily recognizable as spherical, light-brown nodules by
macroscopy. Histology (H&E stain) showed round nodular formation of atypical liver cells compressing adjacent liver parenchyma consistent with murine liver
cell adenoma.
←
119
Downloaded from https://academic.oup.com/carcin/article-abstract/35/1/114/360222
by guest
on 06 November 2017

F.H.T.Duong etal.
conrmed these data on the protein level and showed that PP2Ac
protein is more signicantly upregulated in tumorous tissues than in
non-tumoral surrounding tissues both from non-CHC-related HCC
samples (Figure6B). This upregulation was not visible in CHC-related
HCC tissues because of the elevated expression level of PP2Ac in
non-tumorous liver parenchyma in patients with CHC compared with
control samples from patients with other liver diseases (30). Because
of this increased expression of PP2Ac in non-tumorous tissues, we
excluded CHC patients from the overall analysis. We then analyzed
the expression of Noxa and PUMA, two p53-dependent genes that
were altered by PP2A upregulation in vitro. Our data showed no sig-
nicant difference of Noxa expression between tumorous and non-
tumorous samples (Figure6D). However, we observed a signicant
reduction of PUMA expression in malignant versus non-tumorous
tissue (Figure6C).
Discussion
We have reported previously that HCV-induced PP2Ac upregulation
impairs epigenetic changes on histone H4 and inhibits DNA damage
repair machinery (4,7). In the present study, we show that PP2Ac
also inhibits p53. As a consequence, B6HCV mice that overexpress
PP2Ac in the liver have an increased susceptibility for HCC. We
realize that our HCV-mediated PP2Ac overexpression mouse is an
indirect in vivo model to experimentally explore the consequences of
PP2Ac overexpression. Atransgenic mouse model of PP2Ac overex-
pression would be preferable, but no such animal model is available.
However, given our ndings in cells with overexpression and silenc-
ing of PP2Ac, and the results in regard to serine phosphorylation of
p53 and p53 target gene induction in the B6HCV animal model, we
conclude that PP2A-induced dysregulation of p53 is an important
mechanism of HCC carcinogenesis. Interestingly, in several experi-
mental systems, PP2A has tumor suppressor functions (31). In our
model, PP2A seems to predispose cells to transformation by altering
p53 functions. There have been other reports of the procarcinogenic
role of PP2A, notably in testicular germ cell tumors and for T-cell
leukemia cells (32,33).
DEN is a well-known liver carcinogen (34). DEN is usually
injected at doses 50–200 µg/g of body weight to rats (35–37) and
at doses 75–90 µg/g of body weight to mice (38,39) to induce liver
tumors. Some studies have used DEN at 5 µg/g of body weight
associated with a second injection of a liver tumor promoter, 2
weeks after DEN administration (40). At the dosages of 75–90
µg/g of body weight, mice develop liver tumors by 40 weeks.
Therefore, in order to test if HCV-mediated PP2Ac overexpres-
sion could predispose mice to liver tumors, we have deliberately
injected a single low dose of 5 µg/g of DEN. Our results show that
B6HCV had a higher incidence of liver tumor formation than the
control mice, presumably caused by an initial higher expression
of PP2Ac that facilitates DNA damage and inhibits p53-mediated
apoptosis.
The proapoptotic function of the tumor suppressor p53 is con-
trolled by numerous posttranslational modications on several resi-
dues. We have focused on some specic and characterized residues.
It has been shown that PP2A inhibition or knocking down using
pharmacological inhibitors or small interfering RNA, respectively,
enhanced phosphorylation on serine 46 and induced apoptosis (41).
Additionally, using PP2A inhibitors, it has been demonstrated that
PP2A dephosphorylates serine 37 and negatively controls p53 tran-
scriptional activity (19). In line with these previous observations, we
show that PP2Ac overexpression impairs phosphorylation of p53 on
serine 37 and 46 but not on serine 15. Our observation that etoposide-
induced serine 15 phosphorylation is reduced only in UHCV57.3
cells is intriguing. However, it has been reported previously that
phosphorylation on serine 15 can substantially be suppressed by a
high expression of HCV core in HepG2 cells (42). Therefore, we
hypothesize that the impairment of phosphorylation on serine 15
might be caused by a high expression of the core in these HCV-
inducible cells independently of PP2Ac upregulation. Similarly, the
reduction of serine 18 phosphorylation, a homolog of serine 15 in
human, in HCV transgenic mice was presumably due to an elevated
expression of HCV core in these animals. Surprisingly, we did not
observe an impaired serine 15 phosphorylation in HCV-infected
Huh7 cells. The discrepancy could be caused by a more restricted
subcellular location or a lower expression level of HCV core protein
during HCV replication in the infectious cell culture system com-
pared with UHCV57.3 cells. Further analyses are required to clarify
thisissue.
It has been shown that p53 serine residues can be phosphoryl-
ated by a number of kinases, amongst them ATM, DNA-PK and
p38MAPK (41,43,44). Therefore, PP2A might decrease serine 37
and 46 phosphorylation by regulating one or more of these kinases.
However, the specic dephosphorylation of p53 serine 37 and 46
observed in the in vitro dephosphorylation assays (Supplementary
Figure S2, available at Carcinogenesis Online) provides evidence
for a direct mechanism where the phosphatase enzymatic activ-
ity of PP2A is responsible for the dephosphorylation of serine 37
and46.
We have reported previously that PP2A catalytic subunit is
overexpressed in liver biopsies from patients chronically infected with
Fig.6. PP2Ac is highly expressed in malignant tissues from non-CHC-
related HCC and negatively correlates with PUMA expression. Quantitative
reverse transcription–PCR was performed on paired liver biopsies (CT =
non-malignant surrounding tissues; TU = malignant tissues) from CHC
(n = 7) and non-CHC-derived HCC (n=31). (A) PP2Ac expression
level was determined in CHC and non-CHC samples. (B) PP2Ac protein
expression level was evaluated by immunoblotting from non-CHC samples.
(C) PUMA and (D) Noxa expression levels were measured in non-CHC
samples. Results are shown as box plots, and statistical analysis was
performed using Mann–Whitney’s test.
120
Downloaded from https://academic.oup.com/carcin/article-abstract/35/1/114/360222
by guest
on 06 November 2017

PP2Ac upregulation increases DEN-mediated liver tumors
HCV. Because our in vitro and in vivo data demonstrated that PP2A
overexpression leads to impairment of p53-induced apoptosis and
correlates with more HCC foci formation in the presence of mutagenic
agents, we expected to measure a signicantly higher expression of
PP2Ac in tumorous biopsies than in non-tumorous biopsies from
CHC-related HCC. Because of the elevated expression level of PP2Ac
in non-tumorous liver parenchyma in patients with CHC compared
with control samples from patients with other liver diseases (30),
we were not able to measure a difference in the expression level
between non-tumorous versus tumorous tissues. However, analysis
of paired biopsies from non-CHC-derived HCC clearly showed an
overexpression of PP2Ac in tumorous versus non-tumorous samples.
The transcription factor, cyclic adenosine monophosphate response
element-binding protein (CREB), is overexpressed or activated in
several malignancies (45). For instance, total and active CREB is
enhanced in HCC samples versus normal liver samples in a rat model of
HCC (46). Furthermore, we have reported that endoplasmic reticulum
stress response induces PP2Ac upregulation via CREB activation (5).
Thus, we believe that the increased expression of PP2Ac in tumorous
tissues from non-CHC related HCC is presumably caused by CREB
transcriptional activity. Taken together, our data show an elevated
expression of PP2Ac in tumorous samples and a negative correlation
with PUMA, suggesting that PP2Ac upregulation provides favorable
conditions for carcinogenesis.
In conclusion, we demonstrated that PP2Ac overexpression par-
ticipates in tumor establishment and maintenance by altering the
apoptotic cell death program. Importantly, this is the rst report that
associates HCV-induced PP2Ac overexpression with liver tumors.
Supplementary material
Supplementary Figures S1–S3 can be found at http://carcin.
oxfordjournals.org/
Funding
Schweizerische Krebsliga (KLS-02522-02-2010); Swiss National
Science Foundation (320030_130243); Forschungsfonds der
Universität Basel (DMS2125).
Conict of Interest Statement: None declared.
References
1. Bosch,F.X. et al. (2004) Primary liver cancer: worldwide incidence and
trends. Gastroenterology, 127 (5 suppl. 1), S5–S16.
2. McGivern,D.R. etal. (2011) Virus-specic mechanisms of carcinogenesis
in hepatitis C virus associated liver cancer. Oncogene, 30, 1969–1983.
3. Moriya,K. etal. (1998) The core protein of hepatitis C virus induces hepa-
tocellular carcinoma in transgenic mice. Nat. Med., 4, 1065–1067.
4. Duong,F.H. et al. (2004) Hepatitis C virus inhibits interferon signaling
through up-regulation of protein phosphatase 2A. Gastroenterology, 126,
263–277.
5. Christen,V. etal. (2007) Activation of endoplasmic reticulum stress response
by hepatitis viruses up-regulates protein phosphatase 2A. Hepatology, 46,
558–565.
6. Millward,T.A. etal. (1999) Regulation of protein kinase cascades by pro-
tein phosphatase 2A. Trends Biochem. Sci., 24, 186–191.
7. Duong,F.H. etal. (2010) Hepatitis C virus-induced up-regulation of pro-
tein phosphatase 2A inhibits histone modication and DNA damage repair.
Hepatology, 51, 741–751.
8. Branzei,D. et al. (2008) Regulation of DNA repair throughout the cell
cycle. Nat. Rev. Mol. Cell Biol., 9, 297–308.
9. Saretzki,G. (2010) Cellular senescence in the development and treatment
of cancer. Curr. Pharm. Des., 16, 79–100.
10. Van Hoof,C. etal. (2003) Phosphatases in apoptosis: to be or not to be,
PP2A is in the heart of the question. Biochim. Biophys. Acta, 1640, 97–104.
11. Okamoto,K. et al. (2002) Cyclin G recruits PP2A to dephosphorylate
Mdm2. Mol. Cell, 9, 761–771.
12. Amaral,J.D. et al. (2009) p53 and the regulation of hepatocyte apoptosis:
implications for disease pathogenesis. Trends Mol. Med., 15, 531–541.
13. Freed-Pastor,W.A. et al. (2012) Mutant p53: one name, many proteins.
Genes Dev., 26, 1268–1286.
14. Mallette,F.A. etal. (2007) The DNA damage signaling pathway connects
oncogenic stress to cellular senescence. Cell Cycle, 6, 1831–1836.
15. Chernov,M.V. etal. (2001) Regulation of ubiquitination and degradation of
p53 in unstressed cells through C-terminal phosphorylation. J. Biol. Chem.,
276, 31819–31824.
16. Puca,R. et al. (2010) Regulation of p53 activity by HIPK2: molecu-
lar mechanisms and therapeutical implications in human cancer cells.
Oncogene, 29, 4378–4387.
17. Lavin,M.F. etal. (2006) The complexity of p53 stabilization and activation.
Cell Death Differ., 13, 941–950.
18. Ichwan,S.J. et al. (2006) Defect in serine 46 phosphorylation of p53 con-
tributes to acquisition of p53 resistance in oral squamous cell carcinoma
cells. Oncogene, 25, 1216–1224.
19. Li,D.W. et al. (2006) Protein serine/threonine phosphatase-1 dephospho-
rylates p53 at Ser-15 and Ser-37 to modulate its transcriptional and apop-
totic activities. Oncogene, 25, 3006–3022.
20. Vogelstein,B. et al. (2000) Surng the p53 network. Nature, 408,
307–310.
21. Vousden,K.H. etal. (2007) p53 in health and disease. Nat. Rev. Mol. Cell
Biol., 8, 275–283.
22. Sherr,C.J. etal. (2002) The RB and p53 pathways in cancer. Cancer Cell,
2, 103–112.
23. Moradpour,D. etal. (1998) Continuous human cell lines inducibly express-
ing hepatitis C virus structural and nonstructural proteins. Hepatology, 28,
192–201.
24. Pietschmann,T. et al. (2006) Construction and characterization of infec-
tious intragenotypic and intergenotypic hepatitis C virus chimeras. Proc.
Natl Acad. Sci. USA, 103, 7408–7413.
25. Blindenbacher,A. et al. (2003) Expression of hepatitis c virus pro-
teins inhibits interferon alpha signaling in the liver of transgenic mice.
Gastroenterology, 124, 1465–1475.
26. Sarasin-Filipowicz,M. et al. (2009) Alpha interferon induces long-lasting
refractoriness of JAK-STAT signaling in the mouse liver through induction
of USP18/UBP43. Mol. Cell. Biol., 29, 4841–4851.
27. Thoolen,B. et al. (2010) Proliferative and nonproliferative lesions of
the rat and mouse hepatobiliary system. Toxicol. Pathol., 38 (suppl. 7),
5S–81S.
28. Chao,E.C. etal. (2006) Molecular models for the tissue specicity of
DNA mismatch repair-decient carcinogenesis. Nucleic Acids Res., 34,
840–852.
29. Jang,J.J. et al. (1992) Progressive atypia in spontaneous and
N-nitrosodiethylamine-induced hepatocellular adenomas of C3H/HeNCr
mice. Carcinogenesis, 13, 1541–1547.
30. Duong,F.H. etal. (2006) S-Adenosylmethionine and betaine correct hepa-
titis C virus induced inhibition of interferon signaling in vitro. Hepatology,
43, 796–806.
31. Mumby,M. (2007) PP2A: unveiling a reluctant tumor suppressor. Cell, 130,
21–24.
32. Boudreau,R.T. et al. (2007) Apoptosis induced by protein phosphatase
2A (PP2A) inhibition in T leukemia cells is negatively regulated by
PP2A-associated p38 mitogen-activated protein kinase. Cell. Signal., 19,
139–151.
33. Schweyer,S. etal. (2007) Expression and function of protein phosphatase
PP2A in malignant testicular germ cell tumours. J. Pathol., 213, 72–81.
34. Lim,I.K. (2003) Spectrum of molecular changes during hepatocarcino-
genesis induced by DEN and other chemicals in Fisher 344 male rats
[Mechanisms of Ageing and Development 123 (2002) 1665-1680]. Mech.
Ageing Dev., 124, 697–708.
35. Mukherjee,B. etal. (2005) Characterization of insulin-like-growth factor
II (IGF II) mRNA positive hepatic altered foci and IGF II expression in
hepatocellular carcinoma during diethylnitrosamine-induced hepatocar-
cinogenesis in rats. J. Carcinog., 4, 12.
36. Pascale,R.M. etal. (2005) Role of HSP90, CDC37, and CRM1 as modula-
tors of P16(INK4A) activity in rat liver carcinogenesis and human liver
cancer. Hepatology, 42, 1310–1319.
37. Schiffer,E. etal. (2005) Getinib, an EGFR inhibitor, prevents hepatocel-
lular carcinoma development in the rat liver with cirrhosis. Hepatology, 41,
307–314.
38. Yamamoto,Y. etal. (2004) The orphan nuclear receptor constitutive active/
androstane receptor is essential for liver tumor promotion by phenobarbital
in mice. Cancer Res., 64, 7197–7200.
39. Yoshiji,H. etal. (2004) Halting the interaction between vascular endothe-
lial growth factor and its receptors attenuates liver carcinogenesis in mice.
Hepatology, 39, 1517–1524.
121
Downloaded from https://academic.oup.com/carcin/article-abstract/35/1/114/360222
by guest
on 06 November 2017

F.H.T.Duong etal.
40. Kalinichenko,V.V. etal. (2004) Foxm1b transcription factor is essential for
development of hepatocellular carcinomas and is negatively regulated by
the p19ARF tumor suppressor. Genes Dev., 18, 830–850.
41. Saito,S. etal. (2002) ATM mediates phosphorylation at multiple p53 sites, includ-
ing Ser(46), in response to ionizing radiation. J. Biol. Chem., 277, 12491–12494.
42. Kao,C.F. et al. (2004) Modulation of p53 transcription regulatory activ-
ity and post-translational modication by hepatitis C virus core protein.
Oncogene, 23, 2472–2483.
43. Tibbetts,R.S. etal. (1999) A role for ATR in the DNA damage-induced
phosphorylation of p53. Genes Dev., 13, 152–157.
44. Perfettini,J.L. etal. (2005) Essential role of p53 phosphorylation by p38
MAPK in apoptosis induction by the HIV-1 envelope. J. Exp. Med., 201,
279–289.
45. Conkright,M.D. etal. (2005) CREB: the unindicted cancer co-conspirator.
Trends Cell Biol., 15, 457–459.
46. Kovach,S.J. et al. (2006) Role of cyclic-AMP responsive element binding
(CREB) proteins in cell proliferation in a rat model of hepatocellular carci-
noma. J. Cell. Physiol., 206, 411–419.
Received January 7, 2013; revised June 20, 2013; accepted July 7, 2013
122
Downloaded from https://academic.oup.com/carcin/article-abstract/35/1/114/360222
by guest
on 06 November 2017