Molecular mechanisms of sorafenib action in liver cancer cells

Abstract

Sorafenib, a multikinase inhibitor, recently received FDA approval for the treatment of advanced hepatocellular carcinoma (HCC). However, as the clinical application of sorafenib evolves, there is increasing interest in defining the mechanisms underlying its anti-tumor activity. Considering that this specific inhibitor could target unexpected molecules depending on the biologic context, a precise understanding of its mechanism of action could be critical to maximize its treatment efficacy, while minimizing adverse effects. Two human HCC cell lines (HepG2 and Huh7), carrying different biological and genetic characteristics, were used in this study to examine the intracellular events leading to sorafenib-induced HCC cell-growth inhibition. Sorafenib inhibited cell growth in both cell lines in a dose- and time-dependent manner and significantly altered expression levels of 826 and 2011 transcripts in HepG2 and Huh7 cells, respectively. Genes functionally involved in angiogenesis, apoptosis, transcription regulation, signal transduction, protein biosynthesis and modification were predominantly upregulated, while genes implicated in cell cycle control, DNA replication recombination and repair, cell adhesion, metabolism and transport were mainly downregulated upon treatment. However, each sorafenib-treated HCC cell line displayed specificity in the expression and activity of crucial factors involved in hepatocarcinogenesis. The altered expression of some of these genes was confirmed by semiquantitative and quantitative RT-PCR and by western blotting. Many novel genes emerged from our transcriptomics analysis that had not previously been reported to be effected by sorafenib. Further functional analyses may determine whether these genes can serve as potential molecular targets for more effective anti-HCC strategies.

Full text

© 2012 Landes Bioscience.
Do not distribute.
www.landesbioscience.com Cell Cycle 2843
Cell Cycle 11:15, 2843-2855; August 1, 2012; © 2012 Landes Bioscience
REPORT
REPORT
*Correspondence to: Melchiorre Cervello and Dimcho Bachvarov; Email: cervello@ibim.cnr.it and dimtcho.batchvarov@crhdq.ulaval.ca
Submitted: 04/24/12; Revised: 06/15/12; Accepted: 06/20/12
http://dx.doi.org/10.4161/cc.21193
Introduction
Hepatocellular carcinoma (HCC) is the fifth most frequent
cancer and is currently the third major cause of cancer-related
deaths.
1,2
Although recent progress in diagnostic and treatment
technologies has improved survival, the long-term survival of
HCC patients remains dismal due to the lack of adequate thera-
pies. Conventional chemotherapies are generally ineffective;
hence, the development of novel agents to enhance the effective-
ness of treatment is mandatory.
Molecular targeted therapy, which acts on specific deregulated
signal transduction pathways, has shown promise as a treatment
for advanced HCC.
3,4
Recently, the FDA and EMEA approved
a new drug, sorafenib (Nexavar
®
, BAY43–9006), for the treat-
ment of patients with advanced HCC. Sorafenib is the first
oral multi-kinase inhibitor that targets Raf kinases to be devel-
oped. The Raf proteins are integral components of the Ras/Raf/
Sorafenib, a multikinase inhibitor, recently received FDA approval for the treatment of advanced hepatocellular carcinoma
(HCC). However, as the clinical application of sorafenib evolves, there is increasing interest in dening the mechanisms
underlying its antitumor activity. Considering that this specic inhibitor could target unexpected molecules depending
on the biologic context, a precise understanding of its mechanism of action could be critical to maximize its treatment
ecacy, while minimizing adverse eects. Two human HCC cell lines (HepG2 and Huh7), carrying dierent biological
and genetic characteristics, were used in this study to examine the intracellular events leading to sorafenib-induced
HCC cell-growth inhibition. Sorafenib inhibited cell growth in both cell lines in a dose- and time-dependent manner and
signicantly altered expression levels of 826 and 2011 transcripts in HepG2 and Huh7 cells, respectively. Genes functionally
involved in angiogenesis, apoptosis, transcription regulation, signal transduction, protein biosynthesis and modication
were predominantly upregulated, while genes implicated in cell cycle control, DNA replication recombination and repair,
cell adhesion, metabolism and transport were mainly downregulated upon treatment. However, each sorafenib-treated
HCC cell line displayed specicity in the expression and activity of crucial factors involved in hepatocarcinogenesis. The
altered expression of some of these genes was conrmed by semiquantitative and quantitative RT-PCR and by western
blotting. Many novel genes emerged from our transcriptomics analysis that had not previously been reported to be
eected by sorafenib. Further functional analyses may determine whether these genes can serve as potential molecular
targets for more eective anti-HCC strategies.
Molecular mechanisms of sorafenib action
in liver cancer cells
Melchiorre Cervello,
1,†,
* Dimcho Bachvarov,
2,3,†,
* Nadia Lampiasi,
1
Antonella Cusimano,
1
Antonina Azzolina,
1
James A. McCubrey
4
and Giuseppe Montalto
1,5
1
Institute of Biomedicine and Molecular Immunology “Alberto Monroy”; National Research Council (CNR); Palermo, Italy;
2
Cancer Research Centre; Hôpital L’Hotel-Dieu de
Québec; Centre Hospitalier Universitaire de Québec; Québec, QC Canada;
3
Department of Molecular Medicine; Faculty of Medicine; Laval University; Québec, QC Canada;
4
Department of Microbiology and Immunology; Brody School of Medicine at East Carolina University; Greenville, NC USA;
5
Department of Internal Medicine and Specialties;
University of Palermo; Palermo, Italy
†
These authors contributed equally to this work.
Keywords: sorafenib, HCC, mini-chromosome maintenance genes, Dickkopf1, Harakiri, Acheron/LARP6, YAP1, cell cycle,
microarray, global gene expression analysis
mitogen-activated protein (MAP)/extracellular signal-regulated
kinase (ERK) kinase (MEK)/ERK (Ras/Raf/MEK/ERK) sig-
naling cascade. In addition, sorafenib inhibits the activity of
several tyrosine kinases involved in tumor angiogenesis and pro-
gression, including VEGFR-2/3, PDGFR-β, Flt3 and c-Kit.
4-6
The molecular mechanism(s) by which sorafenib exerts its
antitumor activity has not been fully elucidated, and both Raf/
MEK/ERK-dependent or -independent mechanisms have been
identified.
7-9
Therefore, considering that sorafenib could target
unexpected molecules depending on the biological context, a
precise understanding of its mechanism of action is critical to
maximize its treatment efficacy while minimizing adverse effects.
Herein, we investigated the molecular mechanisms of
sorafenib-mediated cytotoxic/anti-proliferative activity in two
human HCC cell lines (HepG2 and Huh7), which differ con-
siderably in their biological and genetic characteristics and cor-
respondingly display significant differences in Raf/MEK/ERK

© 2012 Landes Bioscience.
Do not distribute.
2844 Cell Cycle Volume 11 Issue 15
downregulation of anti-apoptotic Bcl-2 family members Mcl-1
and survivin, and that this occurs mainly through a MEK/ERK-
independent mechanism.
8,14-16
Therefore, Mcl-1 and survivin pro-
tein levels were determined. Both these two major anti-apoptotic
proteins were considerably downregulated by sorafenib (Fig. 2A).
Analysis of signaling pathways affected by sorafenib treat-
ments. As discussed previously, sorafenib inhibits multiple
kinases, including Raf-1. Activated Raf causes the phosphoryla-
tion and activation of MEK1/MEK2, which, in turn, phosphory-
late and activate ERK1/ERK2. However, our results indicated
that sorafenib might exert its effects on HCC cells independently
of Raf/MEK/ERK inhibition. Indeed, there are substantial data
showing that sorafenib antitumor activity does not fully correlate
with the inhibition of ERK phosphorylation. To further inves-
tigate the mechanisms responsible for the cell-killing activity of
sorafenib, we analyzed the effects of sorafenib treatment of HCC
cells on the expression of MEK/ERK. Our data demonstrated
that sorafenib downregulated p-MEK and p-ERK in Huh7 cells.
However, it did not significantly affect the weak basal p-ERK
expression in HepG2 cells (Fig. 2B), suggesting that sorafenib
might not act by targeting the MEK/ERK pathway in these cells.
Therefore, we analyzed some additional signaling path-
ways known to be involved in liver carcinogenesis. The PI3K/
Akt pathway is one of the key pathways in HCC, its activation
inducing cell proliferation and increasing survival.
4,17
However,
sorafenib neither changed phospho-Akt (Ser 473) nor Akt levels
in either of the cell lines (Fig. 2B); these data are in agreement
with previous findings obtained by others in glioblastoma cells.
18
The Wnt/β-catenin pathway is also of major significance in the
pathophysiology of HCC.
4
Treatment with sorafenib in a dose-
dependent manner inhibited the expression of the wild type (wt)
form of β-catenin but not of the mutated form (mut) present in
HepG2 cells, while it has only a modest effect on the expression
of wt β-catenin in Huh7 cells (Fig. 2B).
Transcriptomics analysis identifies changes in gene expres-
sion both in common and unique to HepG2 and Huh7 cells
following sorafenib treatment. To identify new potential mech-
anisms of action of sorafenib, its effects on global gene expression
in both cell lines were investigated and compared using DNA
microarray technology. Agilent 44K Human Whole Genome
Oligonucleotide Microarrays (containing ~44,000 genes) were
used to identify global gene expression changes in the HCC cell
lines, following treatment with 7.5 μM sorafenib for 48 h. The
concentration was empirically estimated as the maximal drug
concentration that does not cause a considerable reduction in cell
viability (less than 20–30%) and/or changes in cell morphology
pathway activity. In particular, we examined the intracellular
events leading to sorafenib-induced apoptosis and cell growth
inhibition. Further, in order to identify new potential mecha-
nisms of action of sorafenib, the changes were examined both in
proteins by immunoblotting and in gene expression using DNA
microarray technology.
Results
Basal status of MEK/ERK signaling in HCC cells. We first
characterized the basal profile of MEK/ERK pathway activity in
the two HCC cell lines HepG2 and Huh7. These two cell lines
display different characteristics, including differentiation, bio-
logical behavior and genetic defects.
10
Expression of the unphos-
phorylated and phosphorylated forms of MEK and ERK1/2 was
evaluated by western blotting. Phospho-MEK (p-MEK) and
phospho-ERK1/2 (p-ERK1/2) were readily detected in Huh7
cells, whereas only low levels were observed in HepG2 cells
(Fig. 1A).
Sorafenib reduces cell viability, clonogenic survival and
induces apoptosis in HCC cells. Given that steady-state plasma
levels of sorafenib of up to 10 mg/l (16 μM) can be achieved by
oral administration,
11,12
we incubated the two HCC cell lines for
24 and 48 h with sorafenib at concentrations ranging from 1 to
20 μM and assessed the effects of sorafenib on the viability of
the two human HCC cell lines using the MTS assay. As shown
in Figure 1B, a substantial dose-dependent decrease in cell via-
bility was seen in the two cell lines tested, which was compat-
ible with pharmacological plasma levels.
13
After 24 and 48 h of
exposure to the compounds, the IC
50
values were 19.5 ± 1.4 and
12.0 ± 3.1 μM in HepG2 cells, 15.5 ± 4.4 and 11.3 ± 1.4 μM in
Huh7 cells, respectively. Therefore, these data indicated that the
two cell lines have almost the same sensitivity to the inhibitor.
However, since phospho-MEK and phospho-ERK1/2 expression
was barely detectable in HepG2 cells (Fig. 1A), the sorafenib-
mediated growth-inhibitory activity would appear to be indepen-
dent of MEK/ERK pathway inactivation in these cells.
The cytotoxic effects of sorafenib were further confirmed
using a clonogenic assay (Fig. 1D). Cells were treated for 2 d with
or without compounds, the medium was aspirated, and they were
then washed with inhibitor-free medium. Cells were allowed to
grow for an additional 14 d. There was a dose-dependent decrease
in colony-forming ability due to sorafenib treatment in both cell
lines.
With regard to apoptosis, several studies have shown that
treatment with sorafenib is frequently associated with the
Figure 1 (See opposite page). (A) Basal levels of phosphorylated and unphosphorylated MEK and ERK1/2 proteins for the two HCC cell lines.
(B)Sorafenib reduces cell viability in a dose- and time-dependent manner in HepG2 and Huh7 cell lines. Cell viability was determined by MTS assay in
HCC cells after 24 and 48 h of treatment with the indicated concentrations of sorafenib (μM). Data are presented as the percentage of control cells and
are the means ± SD of three separate experiments, each of which was performed in triplicate. (C) Sorafenib reduces cell proliferation in a dose- and
time-dependent manner in HepG2 and Huh7 cell lines. Cell proliferation was determined by estimating the amount of bromodeoxyuridine (BrdU)
incorporation into DNA in HepG2 and Huh7 cells. Cells were treated with the indicated concentrations of sorafenib (μM) for 8 h. Data are presented as
the percentage of control cells and are the means ± SD of three separate experiments, each of which was performed in duplicate. (D) Sorafenib inhib-
ited the clonogenic survival of HCC cells. Cells were plated overnight and exposed to Sorafenib (0–10 μM) for 48 h followed by growth in fresh culture
media for 14 d, as described in Materials and Methods. Surviving colonies were stained (upper panel) and counted (lower panel). Data are expressed as
a percentage of colony in untreated cells and are the mean ± SD of two determinations.

© 2012 Landes Bioscience.
Do not distribute.
www.landesbioscience.com Cell Cycle 2845
selection criteria identified a total of 2,011 differentially expressed
genes (DEGs) in Huh7 and 826 genes in HepG2 cells, respectively
(Fig. 3A). Among these, 881 genes or 399 genes were upregulated
during the treatment period (data not shown). Microarray data
were initially filtered on confidence set at p-value = 0.05, fol-
lowed by filtering onto expression levels ≥ 2 fold. These stringent
Figure 1. For gure legend, see page 2844.

© 2012 Landes Bioscience.
Do not distribute.
2846 Cell Cycle Volume 11 Issue 15
inhibitor increased the expression of genes mostly involved
in apoptosis, immune response and transcription regula-
tion. Genes functionally implicated in the regulation of
cell cycle, cell differentiation, signal transduction and
metabolism were proportionally up- and downregulated,
although genes specifically implicated in lipid metabolism
control displayed predominant induction upon sorafenib
treatment in both HCC cell lines.
Venn diagram analysis (Fig. 3A) prompted us to evalu-
ate whether sorafenib-treated HepG2 and Huh7 cells could
be distinguished on the basis of their gene expression pro-
files. Following filtering on 2-fold signal intensity, we used
a one-way ANOVA parametric test (Welch t-test; variances
not assumed equal) to select discriminatory genes. Indeed,
t-test with a p-value cutoff of 0.05 selected 417 genes with
expression levels differing between the HepG2 and Huh7
cells. Clustering analysis based on the 417-gene list was
performed using the standard condition tree algorithm
provided in GeneSpring and revealed a formation of two
major cluster groups that clearly distinguish HepG2 and
Huh7 cells upon sorafenib treatment (Fig. 3B). Two hun-
dred and six genes from the 417-gene list were upregulated
in HepG2-treated cells compared with Huh7 cells. The
major classifications of these genes included metabolism,
signal transduction, transport and regulation of transcrip-
tion. Genes upregulated in Huh7-treated cells in contrast
to HepG2 cells (211 genes) were mainly involved in metab-
olism, cell proliferation, signal transduction and transport.
The 417-gene list is presented in Table S3.
Pathway and network analyses, based on the 2-fold
p-value = 0.05 DEG lists, were generated using Ingenuity
Pathways Analysis (IPA) software. The IPA comparison
analysis showed that the major functional pathways com-
monly upregulated in both of the HCC cell lines were
involved in cell death, gene expression, cell growth and
proliferation (Fig. 3C), whereas the major functionally
related groups of genes found to be commonly downregu-
lated were involved in cell cycle, DNA replication, recombination
and repair, cellular assembly and organization (Fig. 3D).
A network analysis identified between 12 and 25 highly signif-
icant networks with a score ≥ 3 that were down- or upregulated in
HepG2 and Huh7 cells upon sorafenib treatment. As expected,
the five top-scoring upregulated networks were mainly associated
with functions linked to cell death, immune response, transcrip-
tion regulation (gene expression and developmental regulation)
and metabolism, but excluding lipid metabolism (Table S2A
and B). The top-scoring downregulated networks were mostly
linked with DNA replication recombination and repair, protein
synthesis, lipid metabolism and transport (Table S2C and D).
Common networks, generated by merging the four top-scor-
ing networks that included both down- and upregulated genes
(≤ 2 fold), recognized some functionally related gene nodes
that were similarly modulated in the two HCC cell lines upon
sorafenib treatment (Fig. 4). In particular, a number of gene
nodes implicated in cell cycle control, DNA replication and cell
proliferation were downregulated both in HepG2 cells (CCNB1,
and 1,130 or 427 genes were downregulated in Huh7 and HepG2
cells, respectively. Tables S1 and S2 display the complete list of
the differentially expressed genes (p-value = 0.05 ≥ 2-fold) in the
sorafenib-treated Huh7 and HepG2 cells, respectively. Venn dia-
gram analysis showed that 1,700 and 515 genes were differen-
tially expressed and unique to Huh7 or HepG2 cells, respectively
(Fig. 3A). Of the 1,700 DEGs unique to Huh7 cells, 771 (45%)
were upregulated and 929 (55%) were downregulated. Similarly,
of the 515 DEGs unique to HepG2 cells, 289 (56%) were upreg-
ulated and 226 (44%) downregulated. In contrast, 311 DEGs
were common to both cell lines; of these, 110 were upregulated
(36.5%), and 201 (63.5%) were downregulated (Fig. 3A). All
the genes common to both Huh7 and HepG2 cells that were sig-
nificantly up- or downregulated (p = 0.05) are listed in Table S1.
It should be emphasized that in both cell lines, sorafenib
treatment produced an important reduction in the expression of
genes predominantly associated with cell growth, cell adhesion,
DNA replication, recombination and repair, protein biosynthesis,
transport and angiogenesis (Table S1). On the other hand, the
Figure 2. Eect of sorafenib on apoptosis-related proteins and signaling
molecules. HepG2 and Huh7 cells were treated with sorafenib at the indicated
concentrations for 24 h. After treatment, cells were harvested and lysed and
equal amounts of extracted protein were analyzed for Mcl-1 and survivin (A)
and for β-catenin and total and phospho-MEK, ERK1/2 and AKT expression by
western blotting (B). The data represent two independent experiments with
comparable outcomes.

© 2012 Landes Bioscience.
Do not distribute.
www.landesbioscience.com Cell Cycle 2847
Figure 3. Comparison of common and distinct gene expressions across the various dierentially expressed gene groups in HepG2 and Huh7 cells
upon sorafenib treatment. (A) Venn diagram analysis of common and distinct gene expression in both cell lines. (B) Hierarchical clustering based on
the 417-gene list (2-fold dierence in gene expression; p-value cuto of 0.05), which discriminates HepG2 and Huh7 cells according to their response
to sorafenib treatment. Red signies upregulation, and green, downregulation. (C) IPA comparison of upregulated functional pathways in HepG2 (dark
blue) vs. Huh7 (sky blue) cells. (D) IPA comparison of downregulated functional pathways in HepG2 (dark blue) vs. Huh7 (sky blue) cells. Top functions
that meet a p-value cuto of 0.05 are displayed.

© 2012 Landes Bioscience.
Do not distribute.
2848 Cell Cycle Volume 11 Issue 15
were found to detect similar patterns for the up- and downregu-
lated genes selected for validation.
Validation of microarray findings with western blotting.
Microarray data showed that the gene encoding for survivin
(BIRC5) was downregulated by treatment with sorafenib. As
shown in Figure 2, we had already validated this observation in
both cell lines at the protein level. An intriguing result observed
in the microarray analysis and also validated by qPCR (Table 1
and Fig. 5) was that the expression of the gene encoding a mem-
ber of Dickkopf (DKK) family proteins, DKK1, was inhibited
by sorafenib. DKK members function as secreted Wnt antago-
nists by inhibiting Wnt co-receptors LRP5/6, therefore by block-
ing Wnt signaling DKK members should suppress Wnt-induced
tumor growth. As shown in Figure 6A, we confirmed the same
finding at the protein level by western blotting in both cell lines,
i.e., that sorafenib inhibits DKK1 expression in a dose-depen-
dent manner. To our knowledge, these findings represent the first
demonstration that sorafenib inhibits the expression of DKK1
in HCC.
In addition, and as shown in Figure 4A, IPA analysis revealed
that in HepG2 cells, one of the major upregulated gene nodes is
the c-Jun N-terminal kinase (JNKs) gene family. Therefore, to
better understand the role of JNK, we first analyzed the effects
of sorafenib on JNK signaling by western blotting. As shown
in Figure 6B, sorafenib increased phosphorylation of JNK
in HepG2 cells in a dose-dependent manner. Activated JNK
translocates to the nucleus, where it can regulate the activity of
multiple transcription factors, including c-Jun. Therefore, we
analyzed the effect of sorafenib on c-Jun expression. Both total
and phospho-c-Jun levels were also increased (Fig. 6). To test the
functional significance of JNK/c-Jun activation, we examined
the effects of the JNK-specific inhibitor SP600125 on HepG2
cell viability by MTS assays. SP600125 also inhibited cell growth
in HepG2 cells in a dose-dependent manner when used alone
(data not shown), and it synergized with sorafenib in inhibiting
cell viability in combination experiments, as shown by CalcuSyn
analysis (Table 2A).
A second important node discovered by IPA analysis in
HepG2 cells is the p38
MAPK
node. For this reason, we analyzed
the expression of phospho-p38 and total-p38 by western blotting.
As shown in Figure 6B, treatment with sorafenib increased phos-
phorylation of p38 in HepG2 cells in a dose-dependent man-
ner. Therefore, to study the functional significance of p38
MAPK
activation, we examined the effects of the p38
MAPK
inhibitor
SB203580. SB203580 inhibited cell growth in HepG2 cells in a
dose-dependent manner when used alone (data not shown), and
it antagonized the effect of sorafenib in inhibiting cell viability
in combination experiments, as shown by CalcuSyn analysis
(Table 2B). Overall, the above results indicated that sorafenib-
mediated activation of the JNK signaling pathway had a pro-
tective role against sorafenib-induced cell growth inhibition,
whereas inhibition of p38
MAPK
signaling partially abrogated
sorafenib-induced cell growth inhibition.
As shown in Figure 5, we observed sorafenib-mediated
increased expression levels of genes involved in the ER stress
response. We wanted to determine a potential mechanism for
CDC20, CDC6, CDC25C, Cyclin B, APC, PTTG1, TOP2A,
CTGF, APC, AURKB, some members of the p38 MAPK net-
work and most members of the MCM network) and in Huh7 cells
(CCND1, CDKN2A, Cyclin D1, CDC6, CDC7, CDC25A, E2F,
E2F2, E2F7, EZH2, CTGF, thymidine kinase, ORC1, ORC6L
and the entire MCM gene network). Indeed, it seems that the
highly conserved mini-chromosome maintenance (MCM) genes,
which are involved in the initiation of eukaryotic genome rep-
lication,
19
represent preferential targets for sorafenib inhibition
in HCC cells. Similarly, some apoptosis-related gene nodes were
upregulated both in HepG2 cells (ATM, EIF2C4, EGR1, FRD1)
and Huh7 cells (MAX, RARA, MDM4, CREBBP, GLI1). Gene
nodes specifically downregulated in HepG2 cells included those
implicated in signal transduction (RACGAP1, HSP90AB1),
metabolism (LDL), nucleosome assembly (Histone H1) and
DNA repair (H2AFX) (Fig. 4A). Interestingly, the JUN onco-
gene and some members of its interaction network displayed
upregulation in sorafenib-treated HepG2 cells (see Fig. 4A),
which could be linked to possible compensatory mechanisms fol-
lowing the cytotoxic action of the drug in this HCC cell line. In
Huh7 cells, sorafenib specifically downregulated some members
of the NFκB interaction network, while the NR3C1 gene (asso-
ciated with immune response) was induced following the inhibi-
tor treatment (Fig. 4B).
In addition, IPA indicated that the top two significant canoni-
cal pathways activated by sorafenib are the acute phase response
pathway (i.e., inflammatory response) and hypoxia signaling in
both cell lines (data not shown).
Validation of microarray findings with semi-quantitative
RT-PCR (sqRT-PCR) and quantitative RT-PCR (qRT-PCR).
To validate our microarray results, we arbitrarily selected 11 dif-
ferentially expressed genes following sorafenib treatment and
quantified their expression by sqRT-PCR and, in some cases, by
qRT-PCR in control and in treated cells. All sqRT-PCR and qRT-
PCR analyses were performed in samples previously used for the
microarray experiments and repeated using RNA extracted from
two other, different experiments. Table 1 and Figure 5 show the
gene expression measurements of all the validated genes. The dif-
ferent methods (microarray analysis, sqRT-PCR and qRT-PCR)
Table1. Fold expression of validated genes
HepG2 Huh7
Gene Microarray RT-PCR Microarray RT-PCR
YAP1 4.6 1.2
BIRC5 -2.39 -10 -2.26 -10
DDIT3 5.3 2.7 10.21 2.0
DKK1 -8.7 -2.0 -19.46 -2.5
FABP1 -14.99 -10 -5.43 -3.3
HRK 2.7 7.2 3.43 16.6
IL6R 2.7 1.3 9.36 3.4
LARP6 8.7 6.0 7. 32 1.9
SPP1 -2.43 -1.7 -9.43 -1.25
TRIB3 3.92 1.9 2.38 1.0
VEGF 2.93 2.6 3.93 1.6

© 2012 Landes Bioscience.
Do not distribute.
www.landesbioscience.com Cell Cycle 2849
these increases, therefore we examined
the expression of phospho-eIF2α that
is involved in both protein translation
control and induction of transcription
of pro-apoptotic genes such as CHOP/
DDIT3 and TRIB3. As presented in
Figure 6B, a dose-dependent induc-
tion of phospho-eIF2α was detected in
HepG2 cells after sorafenib treatment,
indicating that sorafenib may sup-
press protein translation and activate
transcription of pro-apoptotic genes in
HepG2 cells.
Finally, common networks
obtained by merging the four top-scor-
ing networks in Huh7 cells (Fig. 4B)
demonstrated sorafenib involvement
in the modulation of different mem-
bers of the NFκB genetic network.
Interestingly, chronic treatment with
sorafenib has been shown to increase
IκBα levels and therefore to inhibit
the activation of NFκB in the brain
of mice with Alzheimer disease.
20
In
addition, sorafenib has recently been
shown to reduce the translocation
and hence activation of NFκB that
follows irinotecan treatment of atypi-
cal teratoid rhabdoid tumor (AT/RT)
cells.
21
These data therefore suggest
that sorafenib may act by inhibiting
Figure 4. Network analysis of dynamic
gene expression in HepG2 (A) and
Huh7(B) cells based on the 2-fold com-
mon gene expression lists obtained
following treatment with sorafenib.
For each cell line, the three top-scoring
networks have been merged and are dis-
played graphically as nodes (genes/gene
products) and edges (the biological rela-
tionships between the nodes). Intensity
of the node color indicates the degree
of up- (red) or down- (green) regulation.
Nodes are displayed using various shapes
that represent the functional class of the
gene product (square, cytokine; vertical
oval, transmembrane receptor; rectangle,
nuclear receptor; diamond, enzyme;
rhomboid, transporter; hexagon, transla-
tion factor; horizontal oval, transcription
factor; circle, other). Edges are displayed
with various labels that describe the
nature of the relationship between the
nodes: — binding only → acts on. The
length of an edge reects the evidence
supporting that node-to-node relation-
ship and edges supported by articles
from the literature are shorter. Dotted
edges represent indirect interaction.

© 2012 Landes Bioscience.
Do not distribute.
2850 Cell Cycle Volume 11 Issue 15
specific for each cell line. This is to be expected, since despite
disparities in Raf/MEK/ERK activity, the two HCC cell lines
also display other significant differences, such as mutations in
β-catenin, N-Ras, p16ink, p53 and other genes.
Some previously reported genes affected by sorafenib and
involved in the regulation of apoptosis, in ER stress response
and DNA damage response, such as survivin (BIRC5), DDIT3
(GADD153; CHOP)
7
and GADD45β
26
were included among the
sorafenib-modulated genes in our study (Fig. 2; Table S1 and S3).
However, most of the validated genes in Table S4 have not been
previously reported as potential targets of sorafenib. Examples
of these novel sorafenib-modulated genes identified in our study
are genes such as La ribonucleoprotein domain family member 6
(LARP6), Harakiri (Hrk), Yes-associated protein 1 (YAP1) and
DKK1.
LARP6, also known as Acheron (Achn), is a RNA bind-
ing protein that binds with high affinity to the 5' stem loop of
mRNAs encoding type I collagen,
27
and its involvement in the
process of liver fibrosis has been suggested.
28
Recently, Acheron/
LARP6 has also been shown to be involved in the development
of human breast cancer, being preferentially upregulated in some
basal-like carcinomas of the breast.
28
Ectopic expression of Achn
drives a number of physiological responses associated with ele-
vated aggressive behavior both in vitro and in an in vivo xeno-
graft model of breast cancer, including proliferation, invasion,
lamellipodia formation, MMP-9 and VEGF expression, tumor
growth and angiogenesis.
29
Achn/LARP6 not only promotes
angiogenesis through its activity in cancer cells, but it also seems
to affect the function of vascular endothelial cells. In fact, recent
findings indicated that Achn may be an effective mediator of vas-
cular endothelial cell proliferation, angiogenesis and wound heal-
ing by regulating the expression of VEGF.
30
However, the role of
Achn/LARP6 in HCC remains to be determined as well as its
role in sorafenib anti-HCC effects.
Harakiri (HRK), also known as death protein 5 (dp5), is a
BH3-only member of the Bcl-2 family that localizes in mem-
branes and induces cell death by interacting with anti-apoptotic
Bcl-2 members, Bcl-X
L
and Bcl-2.
31
HRK expression has been
shown to be linked to ER-stress response and induction of apop-
tosis,
32
and its inactivation by aberrant DNA methylation may
contribute to the development and progression of various human
cancers.
33
Interestingly, the study of Ma et al. suggests that dp5/
the activity of NFκB, although this remains to be demonstrated
in HCC.
Discussion
Sorafenib (Nexavar
®
, BAY43–9006) is the first oral multi-kinase
inhibitor that targets Raf kinases to be developed, and it is the
only drug approved by the FDA and EMEA for the treatment
of patients with advanced HCC. Although sorafenib improves
overall survival and time to progression of patients with advanced
HCC, patient outcome is still poor. A better understanding of the
mechanism(s) of sorafenib action is highly desirable, not only to
improve its efficacy, but also to reduce its additional side effects.
To date, the molecular mechanisms by which sorafenib exerts
its antitumor activity have not been fully elucidated, as both Raf/
MEK/ERK-dependent and -independent mechanisms have been
observed.
7-9
Therefore, this study was designed to determine the
effects of sorafenib on the growth characteristics of two cyto-
genetically distinct human HCC cell line models - HepG2 and
Huh7 - which display different Raf/MEK/ERK pathway activi-
ties. Furthermore, sorafenib-induced alterations in global gene
expression were analyzed using the Agilent 44K Human Whole
Genome Oligonucleotide Microarray and Ingenuity Pathway
Analysis.
To our knowledge, the present work represents the first
attempt to define global changes in gene expression in HCC
cells treated with sorafenib by using high-density microarrays.
The highlight of the present investigation is that sorafenib not
only modulated the expression of hundreds of genes in cells with
constitutive activation of the MEK/ERK pathway (i.e., Huh7
cells), but also in cells with a very low activation of this pathway
(i.e., HepG2 cells), demonstrating that its mode of action is also
due to a MEK/ERK-independent mechanisms. Accordingly, we
found that sorafenib inhibited cell vitality and colony formation
independently of Raf/MEK/ERK activity. Our data support pre-
vious studies demonstrating that sorafenib leads to a reduction
in Mcl-1 and survivin protein levels in cancer cells, including
HCC cells.
8,9,14,18,22-25
However, although these studies have dem-
onstrated that sorafenib reduces Mcl-1 expression, the precise
mechanism responsible for this effect is still unknown. Indeed,
regulation of Mcl-1 expression is complicated as it is regulated
at the transcriptional, translational and post-translational levels
as well as through multiple pathways (e.g., ERK, PI3K/AKT,
Jak/STAT) and by ubiquitination in response to a variety of
agents.
8,9,14,25
Transcriptomics analysis identified a number of genes that
were commonly differentially expressed in both the HCC cell
lines as well as alterations in gene expression patterns that were
Table2A. Synergistic effects comparison between Sorafenib and
SP600125
Sorafenib (µM) SP600125 (µM)
Fraction affected CI
3.75 7.5 0.57 0.82
7.5 15 0.40 0.71
10 20 0.35 0.77
Table2B. Antagonistic effects comparison between Sorafenib and
SB203580
Sorafenib
(µM)
SB203580 (µM)
Fraction affected CI
3.75 11. 25 0.62 0.93
7.5 22.50 0.54 1.45
10 30 0.44 1.19
HepG2 cells were treated with sorafenib and SP600125 or SB203580
at a fixed ratio (1:2 and 1:3, respectively) for 72 h, then cell vitality was
measured by MTT assay. Data were entered into the CalcuSyn pro-
gram (Biosoft) and CI and fraction-affected values were determined.
A CIvalue of < 0.90 to 1.00 indicates synergy, a CI value of 0.90 to 1.10
approximates to additive interactions between the drugs and a CI value
of > 1.10 indicates antagonism (n = 2 independent experiments).

© 2012 Landes Bioscience.
Do not distribute.
www.landesbioscience.com Cell Cycle 2851
patients treated with sorafenib. In the SHARP trial, sorafenib
nearly doubled plasma levels of VEGF, with an increase of 195.7%
in HCC patients after 12 weeks of treatment.
45
Therefore, our in
vitro findings are in agreement with clinical data and support the
fact that HCC cells can produce angiogenic factors in response
to sorafenib exposure. It is now known that in vivo treatment
with angiogenesis inhibitors induces hypoxia in tumor cells,
which, in turn, reduces primary tumor growth but promotes
tumor invasiveness and metastasis.
46
Accumulating evidence has
shown that in HCC, hypoxia not only stimulates proliferation,
47
invasion and drug resistance
48,49
but also induces angiogenesis.
50
Interestingly, our IPA analysis showed that hypoxia signaling is
one of the top canonical pathways activated as a consequence of
sorafenib treatment in both the HCC cell lines. Therefore, one
plausible explanation for the overexpression of VEGF mRNA is
that sorafenib elicits hypoxia response in HCC cells. This may be
of relevance clinically, because, as has been observed in clinical
practice, therapy with angiogenesis inhibitors often does not pro-
long the survival of cancer patients for more than months, since
HRK is a c-Jun target gene, and that its expression is required to
induce neuronal apoptosis.
34
In agreement, we also observed an
increase in the JNK/c-Jun pathway activity after sorafenib treat-
ment, which correlated with the upregulation of HRK mRNA in
HepG2 cells and apoptosis induction. Furthermore, our results
demonstrated that inhibition of the JNK/c-Jun pathway with
selective JNK inhibitor synergized with sorafenib in inhibiting
cell viability.
YAP1, the nuclear effector of the Hippo pathway, is a key
regulator of organ size and a candidate human oncogene, while
YAP has been shown to be an independent prognostic marker
for overall survival and disease-free survival times of HCC
patients,
35
although its precise role is still not completely clear.
36
The functional activity of YAP protein strongly depends on the
phosphorylation state of precise serine and tyrosine (which con-
trol its localization), degradation and interaction with different
proteins.
36,37
Thereby, YAP regulation and cellular context might
have a pivotal role in the choice of its partners and, consequently,
on the final and different outcomes: proliferation/transformation
and death/tumor suppression.
36,37
Indeed, we observed increased
YAP gene expression in sorafenib-treated HepG2 cells. Therefore
additional studies are necessary to clarify the role of the YAP
protein in sorafenib-induced cell growth inhibition in HCC cells.
DKK1 plays a crucial role in head formation in vertebrate
development and is a potent antagonist of Wnt signaling.
38
However, DDK1 was found to be overexpressed in HCC tumors,
which was associated with poor prognosis, thus implying that
DKK1 most probably functions as a potential oncogenic fac-
tor rather than a tumor suppressor/Wnt signaling inhibitor in
HCC.
39
Thus, DKK1 appears to have diverse roles in cancer.
40
Interestingly, DKK1 was one of the most strongly downregulated
genes in both our HCC cell lines. This result, although surpris-
ing, is of interest for its clinical implication. Due to its secretory
nature, DKK1 is present at high levels in the serum of cancer
patients, and this is associated with poor prognosis in various can-
cers, including breast and cervical cancer,
41
esophageal squamous
cell carcinoma
42
and lung cancer.
43
In liver cancer, the study by
Sato et al. showed that in 53.0% (89 of 168) of HCC patients levels
of serum DKK1 protein were significantly higher than in healthy
volunteers, suggesting the great potential of DKK1 as a serum
biomarker for the detection of patients with HCC.
41
Tung et al.
recently demonstrated that serum DKK1 levels showed a step-
wise increase from HBV-associated cirrhotic patients to patients
with early and advanced HBV-associated HCC, and that a high
serum DKK1 protein level was an unfavorable prognostic bio-
marker significantly associated with shorter disease-free survival
rates in HCC patients.
44
In addition, serum DKK1 protein levels
were significantly reduced after liver resection or treatment for
HCC.
43
Therefore, our in vitro findings of a sorafenib-mediated
downregulation of DKK1 protein in HCC cells indirectly suggest
that DKK1 could be potentially used as a serum biomarker for
monitoring the effects of sorafenib treatment in HCC patients.
Strikingly, we found that VEGF mRNA significantly increased
in HCC cells after treatment with sorafenib. Although unex-
pected, this observation is very important for its clinical implica-
tions. In fact, a similar phenomenon has been observed in HCC
Figure 5. Quantitative PCR validation of microarray data. Cells were
treated for 48 h with 7.5 and 15 μM sorafenib and then mRNA expres-
sion levels of CHOP, TRIB3, LARP6 and DKK1 genes were analyzed in
HepG2 (A) and Huh7 (B) cells. The results shown are the means ± SD of
two experiments each performed in triplicate.

© 2012 Landes Bioscience.
Do not distribute.
2852 Cell Cycle Volume 11 Issue 15
maintenance proteins (MCM) that are involved in the
initiation of eukaryotic genome replication were found
to be downregulated. In addition, a number of genes
involved in DNA repair and recombination (XRCC-2,
XRCC-5, FANCA and FANCD2 ) and cell cycle regula-
tion (CDC45L, CDC6 and CDCA5 ) were downregu-
lated by sorafenib. Interestingly, sorafenib has recently
been shown to significantly enhance the anti-prolif-
erative effects of chemoradiation treatment in a dose-
dependent manner by downregulating DNA repair
proteins, particularly ERCC-1 and XRCC-1, in head
and neck squamous cell carcinoma cells.
54
Efficient
DNA repair in cancer cells is an important mechanism
of therapeutic resistance; therefore, taken together, our
observations strongly support the hypothesis that inhi-
bition of the DNA repair pathway by sorafenib would
make HCC cells more sensitive to DNA damaging
agents like chemotherapy drugs and/or radiation treat-
ment. Indeed, although the combination of sorafenib
with the chemotherapeutic drug doxorubicin has not
yet been indicated for routine clinical use, a recent
phase I study conducted in patients with advanced
HCC demonstrated that treatment with sorafenib plus
doxorubicin compared with doxorubicin monotherapy
resulted in a greater median time to progression, lon-
ger overall survival and progression-free survival,
55
thus
establishing the basis for the ongoing phase 3 trial of
sorafenib plus doxorubicin vs. sorafenib alone.
56
In summary, our data identified potential novel
HCC targets of sorafenib and are in agreement with
previous findings indicating that the Raf/MEK/ERK
pathway is not the only target of sorafenib. Future stud-
ies on selected sorafenib-responsive genes identified here
will provide a better understanding of the molecular
details of the events that are responsible for sorafenib’s
antitumor effects. Furthermore, the sorafenib-respon-
sive genes identified in our study might represent a
potential source of novel molecular targets that may be better
suited for the rational design of more effective therapeutic strate-
gies in HCC.
Materials and Methods
Reagents, cell culture, cell viability, clonogenic and prolifera-
tion assays. Sorafenib was purchased from Alexis Biochemical
and was dissolved in dimethyl sulfoxide (DMSO). The human
hepatocellular carcinoma cell lines HepG2 and Huh7 used in
this study were of a low narrow passage number and were main-
tained as previously described.
57
All cells were kept at 5% CO2
and 37°C and routinely screened for the absence of mycoplasma.
Cell viability assays were performed as previously reported.
58
Drug combination studies were designed according to Chou
and Talalay.
59
CalcuSyn software (Biosoft) was used to calculate
the combination index (CI). CI < 1 indicates synergy, a CI of
approximately 1 indicates an additive effect, and CI > 1 indicates
antagonism.
tumors elicit evasive resistance that can variously enable revas-
cularization via alternative proangiogenic signals, increased local
invasiveness and/or enhanced distant metastasis.
51-53
Furthermore, the molecular mechanism responsible for
sorafenib effects on HCC cells was investigated by analyzing pro-
tein changing in the signaling pathways. Of note, we observed
that expression of the oncoprotein β-catenin was reduced by
sorafenib treatment, especially in HepG2 cells. To our knowl-
edge, this is the first study that identifies β -catenin as a poten-
tial target of sorafenib, suggesting that in some HCC cell types,
sorafenib might also act through the inhibition of β-catenin sig-
naling. In addition, exposure of HepG2 cells to sorafenib acti-
vated JNK and p38
MAPK
stress-related pathways, with the two
signaling pathways having opposite roles in sorafenib-induced
cell growth inhibition, from cytoprotective in the case of JNK/c-
Jun activation to cytotoxic on p38
MAPK
activation.
A number of genes involved in DNA replication, recombina-
tion and repair were identified as being downregulated in both cell
lines. Among them, several members of the mini-chromosome
Figure 6. Eect of sorafenib on expression levels of DKK1 and MAPK signaling
molecules. (A) HepG2 and Huh7 cells were treated with sorafenib at the indicated
concentrations for 24 h. After treatment cells were harvested and lysed, and equal
amounts of extracted protein were analyzed for DKK1 expression by western blot-
ting. (B) HepG2 cells were treated with sorafenib at the indicated concentrations
for 24 h. After treatment, cells were harvested and lysed and equal amounts of
extracted protein were analyzed for total and phospho-JNK, c-Jun, and p38 expres-
sion by western blotting. The data represent two independent experiments with
comparable outcomes.

© 2012 Landes Bioscience.
Do not distribute.
www.landesbioscience.com Cell Cycle 2853
ACA GCA ACT GTC-3'), HRK (5'-CTG TGT CCT TGG
AGA AAG CTG-3' and 5'-GTG TTT CTA CGA TCG CTC
CAG-3'), LARP6 (5'-GGA ACA AGC TGG GAT ATG TGA-
3' and 5'-GGT GGT CCT CAT TCA ACT CAA-3'), TRIB3
(5'-GCC ACT GCC TCC CGT CTT G-3' and 5'-GCT GCC
TTG CCC GAG TAT GA-3'), YAP1 (5'-GGC AAA GAC ATC
TTC TGG TCA-3' and 5'-CAT CAT ATT CTG CTG CAC
TGG-3') and β-actin (5'-CAC CAC ACC TTC TAC AAT GAG
C-3' and 5'-AGT ACA GCT ACG AGC AGT TCT TGT T-3').
PCR reactions were performed using the following parameters:
95°C for 5 min, 94°C for 30 sec, 62°C for DKK1, HRK, LARP6,
60°C for β-actin, FABP1, MT2A, 58°C for DDIT3, HMOX1,
52°C for TRIB3 for 30 sec and 72°C for 1 min, followed by a
final extension step of 72°C for 8 min. The number of cycles
was adjusted to allow detection in the linear range. Finally, PCR
products were analyzed by electrophoresis on agarose gel, then
photographed and quantified by densitometric scanning.
Quantitative RT-PCR analysis. The expression of selected
genes was quantified by quantitative real-time PCR (qPCR)
using SYBR Green fluorescence (Qiagen) on StepOnePlus
(Applied Biosystem). QuantiTect Primer Assays for CHOP
(QT00082278), TRIB3 (QT00088543), LARP6 (QT00221445)
and DKK1 (QT00009093) were purchased from QIAGEN and
amplified as recommended. Relative expression was calculated
using the comparative C
t
method. Expression of each gene was
expressed as fold induction compared with control (DMSO)
and corrected with the quantified expression level of β-actin
(QT00095431). The results shown are the means ± SD of two
experiments, each performed in triplicate.
Disclosure of Potential Conflicts of Interest
No potential conflicts of interest were disclosed.
Acknowledgments
This work was supported in part by grants from the Italian
“Ministero dell’Istruzione, dell’Università e della Ricerca
(Ministry for Education, Universities and Research) – MIUR”
PRIN 2008 and FIRB-MERIT n. RBNE08YYBM to M.C.
and G.M.; D.B. is the Head of the Core Genomic Facility at
the CHUQ-Cancer Research Centre, supported by FRSQ-RR
Cancer, all the genomic experiments and data analyses were per-
formed at this facility; M.C. has been also supported in part by
a grant to the CNR from the Italian Ministry of Economy and
Finance for the Project FaReBio di Qualità.
Supplemental Materials
Supplemental materials may be found here:
www.landesbioscience.com/journals/cc/article/21193
The effect of different inhibitor concentrations on cell viability
was also assessed using a clonogenic assay. For this analysis, 1.0–
1.5 x 10
3
cells were plated onto 6-well plates in growth medium,
and after overnight attachment, cells were exposed to various
sorafenib concentrations or vehicle for 48 h. The cells were then
washed with inhibitor-free medium and allowed to grow for 14 d
under inhibitor-free conditions, after which the colonies contain-
ing more than 50 cells were counted. Relative colony formation
was determined by the ratio of the average number of colonies in
sorafenib-treated cells to the average number of colonies in cells
treated with solvent (DMSO). All experiments were performed
in duplicate and repeated twice.
Cell proliferation was determined by estimating the amount
of bromodeoxyuridine (BrdU) incorporation into DNA using a
colorimetric immunoassay (Roche Diagnostics GmbH) as previ-
ously reported.
60
Western blotting analysis. For western blot analysis, whole-
cellular lysates were obtained using RIPA buffer (Cell Signaling
Technologies Inc.) and western blotting were performed as previ-
ously described,10 with primary antibodies raised against sur-
vivin (Abcam Limited), Mcl-1 (Santa Cruz Biotecnology Inc.),
b-catenin (Transduction Laboratories), b-actin and DKK1
(Sigma-Aldrich Srl), phospho-Akt, Akt, phospho-MEK, MEK,
phospho-ERK1/2, ERK1/2, phospho-p38, p38, phospho-JNK,
JNK, phospho-c-Jun, c-Jun, phospho-eIF2alpha and Bcl-2 (Cell
Signaling Technologies Inc.).
Gene expression profiling and data analysis. Gene expres-
sion was analyzed using Agilent 44K Human Whole Genome
Oligonucleotide Microarrays (containing ~44,000 genes), as
previously described.
61-63
All microarray experiments were per-
formed in duplicate, using dye-swap during labeling. GeneSpring
software (Agilent Technologies Inc.) was used to generate lists
of selected genes and for the various statistical and visualization
methods. Network and pathway analyses of the microarray data
were completed using the Ingenuity Pathway Analysis (IPA) soft-
ware (http://www.ingenuity.com).
Semi-quantitative RT-PCR (sqRT-PCR). Validation of
microarray data was performed for selected differentially
expressed genes by sqRT-PCR as previously described.
60
The
β-actin gene was used as the reference gene for normalization.
The following sense and antisense primers were used, respec-
tively, to amplify human DDIT3 (5'-ATG GCA GCT GAG
TCA TTG CC-3' and 5'-TCA TGC TTG GTG CAG ATT
C-3'), DKK1 (5'-CCG AGG AGA AAT TGA GGA AAC-3'
and 5'-CCT TCT TGT CCT TTG GTG TGA-3'), FABP1
(5'-CTC TAT TGC CAC CAT GAG TTT C-3' and 5'-GCT
GAT TCT CTT GAA GAC AAT-3'), HMOX1 (5'-CTC AAA
AAG ATT GCC CAG AAA G-3' and 5'-GCA TAA AGC CCT
References
1. Montalto G, Cervello M, Giannitrapani L, Dantona F,
Terranova A, Castagnetta LA. Epidemiology, risk fac-
tors, and natural history of hepatocellular carcinoma.
Ann N Y Acad Sci 2002; 963:13-20; PMID:12095924;
http://dx.doi.org/10.1111/j.1749-6632.2002.
tb04090.x.
2. Sherman M. Epidemiology of hepatocellular carcinoma.
Oncology 2010; 78(Suppl 1):7-10; PMID:20616577;
http://dx.doi.org/10.1159/000315223.
3. Llovet JM, Bruix J. Molecular targeted therapies in
hepatocellular carcinoma. Hepatology 2008; 48:1312-
27; PMID:18821591; http://dx.doi.org/10.1002/
hep.22506.
4. Cervello M, McCubrey JA, Cusimano A, Lampiasi
N, Azzolina A, Montalto G. Targeted therapy for
hepatocellular carcinoma: novel agents on the horizon.
Oncotarget 2012; 3:236-60; PMID:22470194.
5. Chappell WH, Steelman LS, Long JM, Kempf RC,
Abrams SL, Franklin RA, et al. Ras/Raf/MEK/ERK
and PI3K/PTEN/Akt/mTOR inhibitors: rationale and
importance to inhibiting these pathways in human
health. Oncotarget 2011; 2:135-64; PMID:21411864.
6. Samant RS, Shevde LA. Recent advances in anti-angio-
genic therapy of cancer.Oncotarget 2011; 2:122-34;
PMID:21399234.

© 2012 Landes Bioscience.
Do not distribute.
2854 Cell Cycle Volume 11 Issue 15
34. Ma C, Ying C, Yuan Z, Song B, Li D, Liu Y, et al.
dp5/HRK is a c-Jun target gene and required for
apoptosis induced by potassium deprivation in cerebel-
lar granule neurons. J Biol Chem 2007; 282:30901-
9; PMID:17428807; http://dx.doi.org/10.1074/jbc.
M608694200.
35. Xu MZ, Yao TJ, Lee NP, Ng IO, Chan YT, Zender
L, et al. Yes-associated protein is an independent
prognostic marker in hepatocellular carcinoma. Cancer
2009; 115:4576-85; PMID:19551889; http://dx.doi.
org/10.1002/cncr.24495.
36. Yee KS, O’Neill E. YAP1--friend and foe. Cell Cycle
2010; 9:1447-8; PMID:20372091; http://dx.doi.
org/10.4161/cc.9.8.11480.
37. Bertini E, Oka T, Sudol M, Strano S, Blandino G. YAP:
at the crossroad between transformation and tumor sup-
pression. Cell Cycle 2009; 8:49-57; PMID:19106601;
http://dx.doi.org/10.4161/cc.8.1.7259.
38. Fedi P, Bafico A, Nieto Soria A, Burgess WH, Miki
T, Bottaro DP, et al. Isolation and biochemical char-
acterization of the human Dkk-1 homologue, a novel
inhibitor of mammalian Wnt signaling. J Biol Chem
1999; 274:19465-72; PMID:10383463; http://dx.doi.
org/10.1074/jbc.274.27.19465.
39. Yu B, Yang X, Xu Y, Yao G, Shu H, Lin B, et al.
Elevated expression of DKK1 is associated with cyto-
plasmic/nuclear beta-catenin accumulation and poor
prognosis in hepatocellular carcinomas. J Hepatol
2009; 50:948-57; PMID:19303159; http://dx.doi.
org/10.1016/j.jhep.2008.11.020.
40. Menezes ME, Devine DJ, Shevde LA, Samant RS.
Dickkopf1: a tumor suppressor or metastasis promoter?
Int J Cancer 2012; 130:1477-83; PMID:21953410;
http://dx.doi.org/10.1002/ijc.26449.
41. Sato N, Yamabuki T, Takano A, Koinuma J, Aragaki
M, Masuda K, et al. Wnt inhibitor Dickkopf-1 as a
target for passive cancer immunotherapy. Cancer Res
2010; 70:5326-36; PMID:20551066; http://dx.doi.
org/10.1158/0008-5472.CAN-09-3879.
42. Makino T, Yamasaki M, Takemasa I, Takeno A,
Nakamura Y, Miyata H, et al. Dickkopf-1 expres-
sion as a marker for predicting clinical outcome in
esophageal squamous cell carcinoma. Ann Surg Oncol
2009; 16:2058-64; PMID:19408050; http://dx.doi.
org/10.1245/s10434-009-0476-7.
43. Yamabuki T, Takano A, Hayama S, Ishikawa N, Kato T,
Miyamoto M, et al. Dikkopf-1 as a novel serologic and
prognostic biomarker for lung and esophageal carcino-
mas. Cancer Res 2007; 67:2517-25; PMID:17363569;
http://dx.doi.org/10.1158/0008-5472.CAN-06-3369.
44. Tung EK, Mak CK, Fatima S, Lo RC, Zhao H, Zhang
C, et al. Clinicopathological and prognostic signifi-
cance of serum and tissue Dickkopf-1 levels in human
hepatocellular carcinoma. Liver Int 2011; 31:1494-
504; PMID:21955977; http://dx.doi.org/10.1111/
j.1478-3231.2011.02597.x.
45. Llovet JM, Ricci S, Mazzaferro V, Hilgard P, Gane E,
Blanc JF, et al.; SHARP Investigators Study Group.
Sorafenib in advanced hepatocellular carcinoma. N
Engl J Med 2008; 359:378-90; PMID:18650514;
http://dx.doi.org/10.1056/NEJMoa0708857.
46. Loges S, Mazzone M, Hohensinner P, Carmeliet P.
Silencing or fueling metastasis with VEGF inhibitors:
antiangiogenesis revisited. Cancer Cell 2009; 15:167-
70; PMID:19249675; http://dx.doi.org/10.1016/j.
ccr.2009.02.007.
47. Gwak GY, Yoon JH, Kim KM, Lee HS, Chung
JW, Gores GJ. Hypoxia stimulates proliferation of
human hepatoma cells through the induction of
hexokinase II expression. J Hepatol 2005; 42:358-
64; PMID:15710218; http://dx.doi.org/10.1016/j.
jhep.2004.11.020.
48. Miyoshi A, Kitajima Y, Ide T, Ohtaka K, Nagasawa
H, Uto Y, et al. Hypoxia accelerates cancer invasion of
hepatoma cells by upregulating MMP expression in an
HIF-1alpha-independent manner. Int J Oncol 2006;
29:1533-9; PMID:17088993.
20. Echeverria V, Burgess S, Gamble-George J, Zeitlin
R, Lin X, Cao C, et al. Sorafenib inhibits nuclear
factor kappa B, decreases inducible nitric oxide syn-
thase and cyclooxygenase-2 expression, and restores
working memory in APPswe mice. Neuroscience
2009; 162:1220-31; PMID:19447162; http://dx.doi.
org/10.1016/j.neuroscience.2009.05.019.
21. Jayanthan A, Bernoux D, Bose P, Riabowol K,
Narendran A. Multi-tyrosine kinase inhibitors in pre-
clinical studies for pediatric CNS AT/RT: Evidence
for synergy with Topoisomerase-I inhibition. Cancer
Cell Int 2011; 11:44; PMID:22206574; http://dx.doi.
org/10.1186/1475-2867-11-44.
22. Blechacz BR, Smoot RL, Bronk SF, Werneburg NW,
Sirica AE, Gores GJ. Sorafenib inhibits signal trans-
ducer and activator of transcription-3 signaling in
cholangiocarcinoma cells by activating the phos-
phatase shatterproof 2. Hepatology 2009; 50:1861-
70; PMID:19821497; http://dx.doi.org/10.1002/
hep.23214.
23. Rosato RR, Almenara JA, Coe S, Grant S. The
multikinase inhibitor sorafenib potentiates TRAIL
lethality in human leukemia cells in association with
Mcl-1 and cFLIPL down-regulation. Cancer Res
2007; 67:9490-500; PMID:17909059; http://dx.doi.
org/10.1158/0008-5472.CAN-07-0598.
24. Huber S, Oelsner M, Decker T, zum Büschenfelde
CM, Wagner M, Lutzny G, et al. Sorafenib induces cell
death in chronic lymphocytic leukemia by translational
downregulation of Mcl-1. Leukemia 2011; 25:838-
47; PMID:21293487; http://dx.doi.org/10.1038/
leu.2011.2.
25. Inuzuka H, Fukushima H, Shaik S, Liu P, Lau AW, Wei
W. Mcl-1 ubiquitination and destruction. Oncotarget
2011; 2:239-44; PMID:21608150.
26. Ou DL, Shen YC, Yu SL, Chen KF, Yeh PY, Fan
HH, et al. Induction of DNA damage-inducible
gene GADD45beta contributes to sorafenib-induced
apoptosis in hepatocellular carcinoma cells. Cancer Res
2010; 70:9309-18; PMID:21062976; http://dx.doi.
org/10.1158/0008-5472.CAN-10-1033.
27. Cai L, Fritz D, Stefanovic L, Stefanovic B. Binding of
LARP6 to the conserved 5' stem-loop regulates transla-
tion of mRNAs encoding type I collagen. J Mol Biol
2010; 395:309-26; PMID:19917293; http://dx.doi.
org/10.1016/j.jmb.2009.11.020.
28. Cai L, Fritz D, Stefanovic L, Stefanovic B. Coming
together: liver fibrosis, collagen mRNAs and the
RNA binding protein. Expert Rev Gastroenterol
Hepatol 2009; 3:1-3; PMID:19210106; http://dx.doi.
org/10.1586/17474124.3.1.1.
29. Shao R, Scully SJ Jr., Yan W, Bentley B, Mueller
J, Brown C, et al. The novel lupus antigen relat-
ed protein acheron enhances the development of
human breast cancer. Int J Cancer 2012; 130:544-
54; PMID:21387291; http://dx.doi.org/10.1002/
ijc.26015.
30. Sun R, Chen W, Zhao X, Li T, Song Q. Acheron regu-
lates vascular endothelial proliferation and angiogenesis
together with Id1 during wound healing. Cell Biochem
Funct 2011; 29:636-40; PMID:22139627; http://
dx.doi.org/10.1002/cbf.1799.
31. Inohara N, Ding L, Chen S, Núñez G. Harakiri: A
novel regulator of cell death, encodes a protein that
activates apoptosis and interacts selectively with sur-
vival-promoting proteins Bcl-2 and Bcl-X(L). EMBO
J 1997; 16:1686-94; PMID:9130713; http://dx.doi.
org/10.1093/emboj/16.7.1686.
32. Gurzov EN, Ortis F, Cunha DA, Gosset G, Li M,
Cardozo AK, et al. Signaling by IL-1beta+IFN-gamma
and ER stress converge on DP5/Hrk activation: a novel
mechanism for pancreatic beta-cell apoptosis. Cell
Death Differ 2009; 16:1539-50; PMID:19629134;
http://dx.doi.org/10.1038/cdd.2009.99.
33. Nakamura M, Shimada K, Konishi N. The role
of HRK gene in human cancer. Oncogene 2008;
27(Suppl 1):S105-13; PMID:19641496; http://dx.doi.
org/10.1038/onc.2009.48.
7. Rahmani M, Davis EM, Crabtree TR, Habibi JR,
Nguyen TK, Dent P, et al. The kinase inhibitor
sorafenib induces cell death through a process involving
induction of endoplasmic reticulum stress. Mol Cell
Biol 2007; 27:5499-513; PMID:17548474; http://
dx.doi.org/10.1128/MCB.01080-06.
8. Tai WT, Cheng AL, Shiau CW, Huang HP, Huang
JW, Chen PJ, et al. Signal transducer and activator of
transcription 3 is a major kinase-independent target
of sorafenib in hepatocellular carcinoma. J Hepatol
2011; 55:1041-8; PMID:21354226; http://dx.doi.
org/10.1016/j.jhep.2011.01.047.
9. Ulivi P, Arienti C, Amadori D, Fabbri F, Carloni
S, Tesei A, et al. Role of RAF/MEK/ERK path-
way, p-STAT-3 and Mcl-1 in sorafenib activity in
human pancreatic cancer cell lines. J Cell Physiol
2009; 220:214-21; PMID:19288493; http://dx.doi.
org/10.1002/jcp.21753.
10. Cervello M, Giannitrapani L, Labbozzetta M,
Notarbartolo M, D’Alessandro N, Lampiasi N, et al.
Expression of WISPs and of their novel alternative vari-
ants in human hepatocellular carcinoma cells. Ann N Y
Acad Sci 2004; 1028:432-9; PMID:15650268; http://
dx.doi.org/10.1196/annals.1322.051.
11. Awada A, Hendlisz A, Gil T, Bartholomeus S, Mano M,
de Valeriola D, et al. Phase I safety and pharmacokinet-
ics of BAY 43-9006 administered for 21 days on/7 days
off in patients with advanced, refractory solid tumours.
Br J Cancer 2005; 92:1855-61; PMID:15870716;
http://dx.doi.org/10.1038/sj.bjc.6602584.
12. Strumberg D, Richly H, Hilger RA, Schleucher N,
Korfee S, Tewes M, et al. Phase I clinical and pharmaco-
kinetic study of the Novel Raf kinase and vascular endo-
thelial growth factor receptor inhibitor BAY 43-9006 in
patients with advanced refractory solid tumors. J Clin
Oncol 2005; 23:965-72; PMID:15613696; http://
dx.doi.org/10.1200/JCO.2005.06.124.
13. Minami H, Kawada K, Ebi H, Kitagawa K, Kim YI,
Araki K, et al. Phase I and pharmacokinetic study of
sorafenib, an oral multikinase inhibitor, in Japanese
patients with advanced refractory solid tumors. Cancer
Sci 2008; 99:1492-8; PMID:18477034; http://dx.doi.
org/10.1111/j.1349-7006.2008.00837.x.
14. Yu C, Bruzek LM, Meng XW, Gores GJ, Carter CA,
Kaufmann SH, et al. The role of Mcl-1 downregula-
tion in the proapoptotic activity of the multikinase
inhibitor BAY 43-9006. Oncogene 2005; 24:6861-
9; PMID:16007148; http://dx.doi.org/10.1038/
sj.onc.1208841.
15. Rahmani M, Davis EM, Bauer C, Dent P, Grant
S. Apoptosis induced by the kinase inhibitor BAY
43-9006 in human leukemia cells involves down-
regulation of Mcl-1 through inhibition of translation.
J Biol Chem 2005; 280:35217-27; PMID:16109713;
http://dx.doi.org/10.1074/jbc.M506551200.
16. Liu L, Cao Y, Chen C, Zhang X, McNabola A,
Wilkie D, et al. Sorafenib blocks the RAF/MEK/
ERK pathway, inhibits tumor angiogenesis, and induc-
es tumor cell apoptosis in hepatocellular carcinoma
model PLC/PRF/5. Cancer Res 2006; 66:11851-8;
PMID:17178882; http://dx.doi.org/10.1158/0008-
5472.CAN-06-1377.
17. Steelman LS, Chappell WH, Abrams SL, Kempf RC,
Long J, Laidler P, et al. Roles of the Raf/MEK/ERK
and PI3K/PTEN/Akt/mTOR pathways in controlling
growth and sensitivity to therapy-implications for can-
cer and aging. Aging (Albany NY) 2011; 3:192-222;
PMID:21422497.
18. Siegelin MD, Raskett CM, Gilbert CA, Ross AH,
Altieri DC. Sorafenib exerts anti-glioma activity in
vitro and in vivo. Neurosci Lett 2010; 478:165-70;
PMID:20470863; http://dx.doi.org/10.1016/j.neu-
let.2010.05.009.
19. Noseda M, Karsan A. Notch and minichromosome
maintenance (MCM) proteins: integration of two
ancestral pathways in cell cycle control. Cell Cycle
2006; 5:2704-9; PMID:17172856; http://dx.doi.
org/10.4161/cc.5.23.3515.

© 2012 Landes Bioscience.
Do not distribute.
www.landesbioscience.com Cell Cycle 2855
49. Wu XZ, Xie GR, Chen D. Hypoxia and hepatocellular
carcinoma: The therapeutic target for hepatocellular
carcinoma. J Gastroenterol Hepatol 2007; 22:1178-82;
PMID:17559361; http://dx.doi.org/10.1111/j.1440-
1746.2007.04997.x.
50. Kim KR, Moon HE, Kim KW. Hypoxia-induced
angiogenesis in human hepatocellular carcinoma. J Mol
Med (Berl) 2002; 80:703-14; PMID:12436347; http://
dx.doi.org/10.1007/s00109-002-0380-0.
51. Loges S, Mazzone M, Hohensinner P, Carmeliet P.
Silencing or fueling metastasis with VEGF inhibitors:
antiangiogenesis revisited. Cancer Cell 2009; 15:167-
70; PMID:19249675; http://dx.doi.org/10.1016/j.
ccr.2009.02.007.
52. Pàez-Ribes M, Allen E, Hudock J, Takeda T, Okuyama
H, Viñals F, et al. Antiangiogenic therapy elicits malig-
nant progression of tumors to increased local invasion
and distant metastasis. Cancer Cell 2009; 15:220-
31; PMID:19249680; http://dx.doi.org/10.1016/j.
ccr.2009.01.027.
53. Ebos JM, Lee CR, Cruz-Munoz W, Bjarnason GA,
Christensen JG, Kerbel RS. Accelerated metastasis
after short-term treatment with a potent inhibitor
of tumor angiogenesis. Cancer Cell 2009; 15:232-
9; PMID:19249681; http://dx.doi.org/10.1016/j.
ccr.2009.01.021.
54. Yadav A, Kumar B, Teknos TN, Kumar P. Sorafenib
enhances the antitumor effects of chemoradiation treat-
ment by downregulating ERCC-1 and XRCC-1 DNA
repair proteins. Mol Cancer Ther 2011; 10:1241-51;
PMID:21551262; http://dx.doi.org/10.1158/1535-
7163.MCT-11-0004.
55. Abou-Alfa GK, Johnson P, Knox JJ, Capanu M,
Davidenko I, Lacava J, et al. Doxorubicin plus
sorafenib vs doxorubicin alone in patients with
advanced hepatocellular carcinoma: a randomized trial.
JAMA 2010; 304:2154-60; PMID:21081728; http://
dx.doi.org/10.1001/jama.2010.1672.
56. National Cancer Institute clinical trials Web page.
Phase III randomized study of sorafenib tosylate
with versus without doxorubicin hydrochloride in
patients with locally advanced or metastatic hepa-
tocellular carcinoma. http://www.cancer.gov/search/
ViewClinicalTrials.aspx?cdrid=659348&version=Healt
hProfessional&protocolsearchid=7262229
57. Cervello M, Notarbartolo M, Landino M, Cusimano
A, Virruso L, Montalto G, et al. Downregulation of
wild-type beta-catenin expression by interleukin 6 in
human hepatocarcinoma HepG2 cells: a possible role
in the growth-regulatory effects of the cytokine? Eur
J Cancer 2001; 37:512-9; PMID:11267862; http://
dx.doi.org/10.1016/S0959-8049(00)00421-4.
58. Cusimano A, Azzolina A, Iovanna JL, Bachvarov D,
McCubrey JA, D’Alessandro N, et al. Novel combina-
tion of celecoxib and proteasome inhibitor MG132
provides synergistic antiproliferative and proapop-
totic effects in human liver tumor cells. Cell Cycle
2010; 9:1399-410; PMID:20305374; http://dx.doi.
org/10.4161/cc.9.7.11254.
59. Chou TC, Talalay P. Quantitative analysis of dose-effect
relationships: the combined effects of multiple drugs or
enzyme inhibitors. Adv Enzyme Regul 1984; 22:27-
55; PMID:6382953; http://dx.doi.org/10.1016/0065-
2571(84)90007-4.
60. Lampiasi N, Azzolina A, D’Alessandro N, Umezawa
K, McCubrey JA, Montalto G, et al. Antitumor
effects of dehydroxymethylepoxyquinomicin, a novel
nuclear factor-kappaB inhibitor, in human liver cancer
cells are mediated through a reactive oxygen species-
dependent mechanism. Mol Pharmacol 2009; 76:290-
300; PMID:19461054; http://dx.doi.org/10.1124/
mol.109.055418.
61. Bachvarov D, Bachvarova M, Koumangaye R, Klein J,
Pesquero JB, Neau E, et al. Renal gene expression pro-
filing using kinin B1 and B2 receptor knockout mice
reveals comparable modulation of functionally related
genes. Biol Chem 2006; 387:15-22; PMID:16497160;
http://dx.doi.org/10.1515/BC.2006.004.
62. L’Espérance S, Popa I, Bachvarova M, Plante M,
Patten N, Wu L, et al. Gene expression profiling of
paired ovarian tumors obtained prior to and follow-
ing adjuvant chemotherapy: molecular signatures of
chemoresistant tumors. Int J Oncol 2006; 29:5-24;
PMID:16773180.
63. Cervello M, Bachvarov D, Cusimano A, Sardina F,
Azzolina A, Lampiasi N, et al. COX-2-dependent
and COX-2-independent mode of action of celecoxib
in human liver cancer cells. OMICS 2011; 15:383-
92; PMID:21410330; http://dx.doi.org/10.1089/
omi.2010.0092.