Báo cáo khoa học: Betulinic acid-mediated inhibitory effect on hepatitis B virus by suppression of manganese superoxide dismutase expression pot

Betulinic acid-mediated inhibitory effect on hepatitis B
virus by suppression of manganese superoxide
dismutase expression
Dachun Yao
1,2
, Huawen Li
3
, Yulan Gou
1,4
, Haimou Zhang
5
, Athanasios G. Vlessidis
2
, Haiyan Zhou
4
,
Nicholaos P. Evmiridis
2
and Zhengxiang Liu
1
1 Internal Medicine of Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China
2 Laboratory of Analytical Chemistry, Department of Chemistry, University of Ioannina, Greece
3 Department of Nutrition and Food Hygiene, Guangdong Medical College, China
4 The First Hospital of Wuhan, China
5 School of Life Sciences, Hubei University, Wuhan, China
Hepatitis B virus (HBV) infection is a prevalent health
problem, affecting 350 million people worldwide; it
causes acute and chronic hepatitis, some cases of
which may progress into cirrhosis and hepatocellular
carcinoma [1]. Chronic HBV patients are currently
treated with interferon or some nucleotide analogs,

including lamivudine and adefovir, but the poor suc-
cess and frequent recurrence after cessation of therapy
require new strategies for terminating this viral infec-
tion. Some complementary and alternative medicines,
Keywords
apoptosis; CREB; mitochondrial;
Pulsatilla chinensis; reactive oxygen species
Correspondence
D. Yao and Z. Lin, Internal Medicine of
Tongji Hospital, Tongji Medical College,
Huazhong University of Science and
Technology, Wuhan 430030, China
Fax: +86 27 83662622
Tel: +86 27 83662601
E-mail: ;

(Received 3 December 2008, revised 26
February 2009, accepted 27 February 2009)
doi:10.1111/j.1742-4658.2009.06988.x
The betulinic acid (BetA) purified from Pulsatilla chinensis (PC) has been
found to have selective inhibitory effects on hepatitis B virus (HBV). In
hepatocytes from HBV-transgenic mice, we showed that BetA substantially
inhibited HBV replication by downregulation of manganese superoxide
dismutase (SOD2) expression, with subsequent reactive oxygen species gen-
eration and mitochondrial dysfunction. Also, the HBV X protein (HBx) is
suppressed and translocated into the mitochondria followed by cyto-
chrome c release. Further investigation revealed that SOD2 expression was
suppressed by BetA-induced cAMP-response element-binding protein
dephosphorylation at Ser133, which subsequently prevented SOD2 tran-
scription through the cAMP-response element-binding protein-binding

motif on the SOD2 promoter. SOD2 overexpression abolished the inhibi-
tory effect of BetA on HBV replication, whereas SOD2 knockdown mim-
icked this effect, indicating that BetA-mediated HBV clearance was due to
modulation of the mitochondrial redox balance. This observation was fur-
ther confirmed in HBV-transgenic mice, where both BetA and PC crude
extracts suppressed SOD2 expression, with enhanced reactive oxygen
species generation in liver tissues followed by substantial HBV clearance.
We conclude that BetA from PC could be a good candidate for anti-HBV
drug development.
Abbreviations
BetA, betulinic acid; CREB, cAMP-response element-binding protein; DiOC
6,
3,3¢-dihexiloxadicarbocyanine; HBeAg, hepatitis B external core
antigen; HBsAg, hepatitis B surface antigen; HBV, hepatitis B virus; HBx, hepatitis B virus X protein; MMP, mitochondrial membrane
potential; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide; PC, Pulsatilla chinensis; PKA, protein kinase A; PKD, protein
kinase D; ROS, reactive oxygen species; siCREB, small interfering RNA for cAMP-response element-binding protein; siRNA, small interfering
RNA; siSOD2, small interfering RNA for manganese superoxide dismutase; SOD2, manganese superoxide dismutase; TUNEL,
deoxynucleotidyl transferase dUTP nick end labeling; WT, wild-type.
FEBS Journal 276 (2009) 2599–2614 ª 2009 The Authors Journal compilation ª 2009 FEBS 2599
including some herbs, have been used for centuries to
treat viral hepatitis, but they are still not widely
accepted by conventional medicine, owing to the lack
of mechanisms and purity of herbs [2].
Pulsatilla chinensis (PC) is a traditional Chinese herb
used for the treatment of amoebic diseases, vaginal
trichomoniasis, and bacterial infections, owing to its
antiamoebic, antibacterial and antitrichomonal activi-
ties [3]. Recently, this herb was used for the treatment
of a hepatitis B patient, according to an old recipe in a
specific area of China (Yichang, Hubei), with satisfac-

tory results for HBV clearance. In order to determine
the mechanism of this, about 30 components from PC
were isolated, and each of them was tested for HBV
clearance. The results revealed that the active compo-
nents were betulinic acid (BetA) and its derivatives
[4,5]. BetA, identified as a pentacyclic triterpene, is
widely available from common natural sources and
possesses several biological properties, including anti-
inflammatory, antiviral, antimalarial, and antimicro-
bial, as well as impressive anticancer and anti-HIV
activities [6–8], although the exact mechanism remains
unclear [9,10].
Manganese superoxide dismutase (SOD2) is an anti-
oxidant enzyme located in mitochondria that can scav-
enge superoxide anions (O
2
·
)
) to form hydrogen
peroxide. Suppression of SOD2 expression may lead to
the overgeneration of reactive oxygen species (ROS)
from mitochondria, and this can subsequently trigger
mitochondrial dysfunction and apoptosis. Altered
SOD2 expression is considered to be both beneficial
and detrimental. For instance, overexpression of SOD2
could be protective against ROS-mediated cell damage,
but it may also increase the invasiveness of tumors
and increase the possibility of infection [11,12]. Several
transcription factors, including specificity protein 1
and nuclear factor-jB [13,14], as well as methylation

[15,16], have been studied extensively for the regulation
of SOD2 expression, whereas there are few reports on
the role of cAMP-response element-binding protein
(CREB) in SOD2 expression [17,18]. CREB binds via
its basic leucine zipper domain as a dimer to cAMP
response elements containing the consensus motif
5¢-TGACGTCA-3¢; these are present in the promoters
of many genes in which transcription rates are strongly
regulated by cAMP. CREB stimulates cellular gene
transcription via the protein kinase A (PKA)-mediated
phosphorylation of CREB at Ser133 [19]. Ser133 phos-
phorylation of CREB, in turn, promotes recruitment
of the coactivator paralogs CREB-binding protein and
p300 via a kinase-inducible domain in CREB, which
appears to be sufficient for the induction of cellular
genes [20,21]. On the other hand, inhibition of CREB
phosphorylation or dephosphorylated CREB may be a
negative regulator of CREB-responsive genes [22,23].
In an effort to investigate the mechanism of the
inhibitory effect of BetA on HBV, BetA was isolated
from PC to treat hepatocytes from HBV-transgenic
mice. We found that SOD2 was downregulated by
BetA-induced CREB dephosphorylation at Ser133
through the CREB-binding motif on the SOD2 pro-
moter. SOD2 suppression-mediated ROS generation
subsequently inhibited HBV replication, decreased
HBV X protein (HBx) total level, and translocated
HBx to the mitochondria followed by cytochrome c
release. Overexpression of SOD2 totally abolished the
BetA-mediated HBV-inhibitory effect, whereas SOD2

knockdown mimicked this effect, indicating that the
BetA-induced HBV-inhibitory effect is due to SOD2
suppression and subsequent ROS generation. Further
in vivo experiments with HBV-transgenic mice con-
firmed our hypothesis; we found that BetA or PC
crude extracts achieved significant HBV clearance, with
decreased SOD2 expression and increased ROS genera-
tion in liver tissue. This is the first time that suppres-
sion of SOD2 expression has been found to be the
mechanism by which BetA inhibits HBV replication.
Results
BetA-induced selective cytotoxicity in
HBV-infected hepatocytes
We first examined the cytotoxicity of BetA in wild-
type (WT) and HBV-infected hepatocytes. Different
dosages of BetA were used to treat the cells for 48 h.
The 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazo-
lium bromide (MTT) assay results showed that there
was little effect on WT cells, whereas BetA treatment
caused significant cytotoxicity in HBV-infected cells
(Fig. 1A). Also, the time course results showed that
WT cells were more resistant to BetA-mediated cyto-
toxicity than HBV-infected cells (Fig. 1B). On the
basis of the above observation, we further evaluated
BetA-mediated cell proliferation; as shown in Fig. 1C,
HBV-infected cells showed a higher DNA synthesis
rate than WT cells with a low dose (5 lgÆmL
)1
)of
BetA, whereas with a high dose (15 lgÆmL

)1
), the
DNA synthesis rate of HBV-infected cells was sub-
stantially decreased, but WT cells showed no signifi-
cant decrease. On the other hand, when the BetA dose
was even higher (20 lgÆmL
)1
), the DNA synthesis rate
of HBV-infected cells was substantially inhibited,
whereas no difference was found in WT cells, indi-
cating that BetA-induced cytotoxicity was specific
to HBV-infected cells.
Betulinic acid inhibits hepatitis B virus D. Yao et al.
2600 FEBS Journal 276 (2009) 2599–2614 ª 2009 The Authors Journal compilation ª 2009 FEBS
96 84
72
60
48
36 24 12 0
2
3
3.5
A C
B D
E F
G H
2.5
1.5
2
3

3.5
2.5
1.5
30 25
20 15 10 5 0
Dosages of betulinic acid (µg·mL
–1
)
Betulinic acid exposing time (h)
Cell viability (A/10
6
cells)
Cell viability (A/10
6
cells)
WT
HBV
WT
HBV
0
300
200
100
0
300
200
100
0
300
200

100
0
300
200
100
400
300
200
100
0 µg·mL
–1
10 µg·mL
–1
15 µg·mL
–1
20 µg·mL
–1
[3H]-thymidine incorporation
(% control)
WT
WT HBV
WT
HBV
WT
HBV
WT
HBV
WT
HBV
HBV

*
*
*
*
¶
¶
0
ROS formation (Arbitrary units)
CTL
BetA
*
¶
*
*
Intracellular ATP level
(Arbitrary units)
CTL
BetA
CTL
BetA
CTL
BetA
CTL
BetA
*
¶
Relative ΔΨ
m
*
¶

Caspase-3 activity
(pmol
–1
·min
–1
·mg
–1
)
*
*
0
5
20
15
10
Apoptosis rate (%)
*
*
Fig. 1. BetA-mediated selective effect on HBV-infected hepatocytes. (A) WT or HBV-infected (HBV) hepatocytes were treated with different
doses of BetA for 48 h, and cell viability was measured. (B) Cells were treated with 15 lgÆmL
)1
BetA for different times, and cell viability
was measured. (C) Cells were treated with different doses of BetA as indicated for 48 h, and then incubated with [
3
H]thymidine for 2 h to
measure the inhibitory effect of BetA on cell differentiation by the [
3
H]thymidine incorporation assay. *P < 0.05 versus WT; –P < 0.05
versus 0 lgÆmL
)1

group. (D–H) Cells were treated with 15 lgÆmL
)1
BetA for 48 h, and the related parameters were measured. (D) BetA-
induced ROS generation. (E) Intracellular ATP level. (F) MMP (Dw
m
). (G) Apoptosis rate determined by TUNEL assay. (H) Intracellular
caspase-3 activity. *P < 0.05 versus control (CTL); –P < 0.05 versus WT group.
D. Yao et al. Betulinic acid inhibits hepatitis B virus
FEBS Journal 276 (2009) 2599–2614 ª 2009 The Authors Journal compilation ª 2009 FEBS 2601
BetA-mediated ROS generation and
mitochondrial dysfunction was specific to
HBV-infected hepatocytes
ROS generation was then examined, and the results
are shown in Fig. 1D. BetA substantially induced ROS
generation in HBV-infected hepatocytes, as compared
with WT cells. As BetA inhibited HBV-infected cell
growth with increased ROS generation, we hypothe-
sized that BetA might also specifically affect mitochon-
drial function in those cells. Measurement of
intracellular ATP generation (Fig. 1E) revealed that
BetA treatment substantially decreased intracellular
ATP generation in HBV-infected cells, but showed no
effect on WT cells. In addition, mitochondrial mem-
brane protential (MMP, DW
m
) was substantially
decreased in HBV-infected cells, but no difference was
found in WT cells (Fig. 1F). Finally, the apoptosis
rates determined by terminal deoxynucleotidyl transfer-
ase dUTP nick end labeling (TUNEL) assay (Fig. 1G)

and caspase-3 activity (Fig. 1H) were assessed. The
results showed that BetA substantially increased the
apoptosis rate and caspase-3 activity in HBV-infected
cells as compared with WT cells.
BetA-mediated selective SOD2 suppression in
HBV-infected hepatocytes
In order to clarify the effect of BetA, a microarray
assay after treatment with 15 lgÆmL
)1
BetA for 48 h
was conducted. BetA specifically decreased SOD2
mRNA expression in HBV-infected cells, whereas little
difference was seen in WT cells (data not shown).
Real-time PCR was performed for confirmatory pur-
poses, and suggested that the SOD2 mRNA level was
decreased about  2.4-fold in HBV-infected cells trea-
ted with BetA as compared with the control, but
showed no difference in WT cells (Fig. 2A). Western
blotting to measure the protein level (Fig. 2B) showed
a significant decrease in SOD2 protein in HBV-infected
cells after BetA treatment, but no change in WT cells.
SOD2 enzyme activity (Fig. 2C) decreased significantly
in HBV-infected cells after BetA treatment, whereas
little difference was found in WT cells.
The BetA-mediated SOD2 transcriptional
response element was located at the
CREB-binding site (nucleotide )1335) on the
SOD2 promoter
The mechanism of BetA-mediated SOD2 suppression
was investigated further. To localize the regulatory

elements required for transcriptional suppression of
the SOD2 gene by BetA treatment, progressive 5¢-
promoter deletion constructs, including )2000,
)1500, )1200, )1000, )500, )200, )100, and 0, were
generated (numbered according to Ensembl Tran-
script ID: ENST00000337404). As shown in Fig. 3A,
the )2000 and )1500 constructs showed a decrease
in activity of about 55%, whereas, with other dele-
tions from )1200 to 0, the reporter activity showed
no significant decrease after BetA treatment. These
data indicate that promoter elements between )1500
and )1200 are responsible for BetA-induced tran-
scriptional suppression of the SOD2 promoter. Com-
parison of these sequences with transcription factor
databases (TFSEARCH) revealed several potential
binding motifs, including GATA ()1488), c-Ets
()1377), CREB ()1335) and NRF2 ()1247). The
0
400
A
B
C
300
200
100
0
400
300
200
100

0
400
300
200
100
WT HBV
WT HBV
WT HBV
SOD2 mRNA level by qPCR
(Arbitrary units)
CTL
BetA
CTL
BetA
CTL
BetA
¶
*
SOD2 protein by Western blot
(Arbitrary units)
¶
*
SOD2 protein activity
(Arbitrary units)
¶
*
Fig. 2. BetA-mediated selective SOD2 suppression in HBV-infected
hepatocytes. The 80% confluent WT or HBV-infected cells were
treated with 15 lgÆmL
)1

BetA for 48 h, and SOD2 expression and
activity were measured. (A) mRNA level. (B) Protein level. (C)
SOD2 enzyme activity. *P < 0.05 versus control (CTL); –P < 0.05
versus WT group.
Betulinic acid inhibits hepatitis B virus D. Yao et al.
2602 FEBS Journal 276 (2009) 2599–2614 ª 2009 The Authors Journal compilation ª 2009 FEBS
possible involvement of these motifs in BetA-induced
SOD2 transcriptional suppression was explored using
a series of luciferase constructs with single mutations.
As shown in Fig. 3B, the SOD2 reporter with the
CREB-binding motif single mutation at )1335 from
nucleotides C to T totally abolished the BetA-
induced SOD2 suppression, whereas the mutations in
other motifs did not decrease the effect (data not
shown). This indicates that the CREB motif at
)1335 is required for BetA responsiveness of the
SOD2 promoter. As the CREB-binding motif was
localized to the BetA-responsive element, the effect
of CREB protein on SOD2 reporter activity was
examined. The SOD2 WT reporter (SOD2 )1500)
showed suppression by BetA treatment, overexpres-
sion of CREB in the presence of BetA totally abol-
ished the effect, and CREB knockdown alone [small
interfering RNA (siRNA) for CREB (siCREB)] mim-
icked this effect (Fig. 3C). On the other hand, the
SOD2 mutation reporter [SOD2 )1500 ⁄ )1335(T)]
showed no effect of either BetA, overexpression of
CREB in the presence of BetA, or siCREB alone,
further demonstrating that the BetA-induced SOD2
suppression is regulated by CREB.

BetA-mediated SOD2 suppression is due to
BetA-induced CREB dephosphorylation
We have shown transcriptional activities of SOD2
that responsible to BetA treatment is due to the exis-
tence of CREB-binding elements on SOD2 promoter.
Here, we further confirmed the CREB-binding activ-
ity through chromatin immunoprecipitation analysis,
as shown in Fig. 4A. After immunoprecipitation and
reversal of the crosslinking, the endogenous SOD2
promoter was enriched by real-time PCR amplifica-
tion, using specific primers that cover the CREB-
binding motif. The results showed that the PCR
product was decreased to 47% after BetA treatment
as compared with the control group, and that the
effect was totally abolished by CREB overexpression
in the presence of BetA, whereas CREB knockdown
(siCREB) mimicked the effect. As it is well known
that CREB activity mainly depends on phosphoryla-
tion at Ser133, we next measured the levels of both
CREB protein and CREB protein phosphorylated at
Ser133 (pCREB). As shown in Fig. 4B,C, the total
CREB protein level did not change after BetA treat-
ment as compared with control, whereas the pCREB
level decreased by 42%. On the other hand, over-
expression of CREB in the presence of BetA
increased the CREB level 1.7-fold, but did not
increase the pCREB level, whereas knockdown of
CREB (siCREB) decreased the levels of both CREB
protein and pCREB. Using the above treatment, we
next measured the SOD2 mRNA level (Fig. 4D) and

0
200
100
CTL BetA si CREB
BetA / CREB
SOD2 transcriptional activity
(Arbitrary units)
SOD2–1500
SOD2–1500/ – 1335 (T)
*
*
0
120
A
B
C
100
80
60
40
20
SOD2–2000
SOD2–1500
SOD2–1200
SOD2–1000
SOD2–500
SOD2–200
SOD2–100
SOD2–0
SOD2 transcriptional activity

(Arbitrary units)
CTL
Bet A
*
*
*
#
#
*
*
*
0
BERC
cuL
TCT
G
CAGT
T
CTGCAGT
TCTGTA
G
T
BERC
0
c
u
L
BERC
0
cuL

0
50
100
150
CTL BetA
–1500
–1335
–1335
–1335
–1500
–1500
SOD2–1500
SOD2–1500
Mut–1335(T)
BERC BERCCREB
cuL cuL
+308
TCT
G
CAGT
T
CTGCAGT
TCTGTA
G
T
T
CTGCAGT
TCTGTA
G
T

T
CTGCAGT
TCTGTA
G
T
BERC
c
u
L
BERC
c
u
L
BERC BERC
c
uL
c
uL
Luc
+ 308
BERC
+ 308
cuL
BERC
cuL
BERC BERC
cuL cuL
SOD2 transcriptional activity
(Arbitrary units)
*

TGACGTCT
CREB
CREB
Luc
Luc
Fig. 3. Mapping of the BetA-responsive element on the SOD2 pro-
moter. (A) HBV-infected hepatocytes were transfected with the
indicated SOD2 reporter constructs, and then treated with either
control (CTL) or 15 lgÆmL
)1
BetA for 48 h; the SOD2 reporter
activity was then measured. *P < 0.05 versus CTL in the SOD2–
2000 group; –P < 0.05 versus CTL. (B) The above cells were
transfected with either SOD2–1500 reporter WT construct or
SOD2–1500 single mutant )1335(T); after the treatment as
indicated above, SOD2 reporter activity was measured. *P < 0.05
versus CTL. (C) HBV-infected hepatocytes were transfected with
either SOD2 )1500 or SOD2 )1500 ⁄ )1335(T) single mutant
reporters, and then treated with CTL, 15 lgÆmL
)1
BetA, BetA with
CREB overexpression (BetA ⁄ CREB›) or siCREB for 48 h, and
SOD2 reporter activity was measured. *P < 0.05 versus CTL in
the SOD2 )1500 group.
D. Yao et al. Betulinic acid inhibits hepatitis B virus
FEBS Journal 276 (2009) 2599–2614 ª 2009 The Authors Journal compilation ª 2009 FEBS 2603
protein level (Fig. 4E). The results showed that both
SOD2 protein expression and mRNA expression
were decreased after BetA treatment, and that this
effect was abolished by CREB overexpression in the

presence of BetA, but was mimicked by siCREB.
This indicates that SOD2 expression is regulated by
CREB phosphorylation at Ser133. As we had
already shown that BetA treatment decreased CREB
phosphorylation at Ser133 (Fig. 4C), we next per-
formed in vitro experiments to determine whether
BetA could inhibit CREB phosphorylation directly.
As shown in Fig. 4F, the purified CREB was sub-
stantially phosphorylated at Ser133 in the presence
of PKA, whereas phosphorylation was markedly
inhibited by BetA, indicating that BetA could
directly inhibit CREB phosphorylation, and this
SOD2 mRNA level by qPCR
(Arbitrary units)
*
*
0
150
A
B
C
D
E
F
100
50
0
300
200
200

100
0
300
200
100
100
qPCR on SOD2 promoter by CREB ChIP
(Arbitrary units)
*
*
0
200
100
0
200
100
0
pCREB (Ser133) level by Western blot
(Arbitrary units)
*
*
CREB protein level by Western blot
(Arbitrary units)
*
*
SOD2 protein level by Western blot
(Arbitrary units)
*
*
pCREB level by Western blot

(Arbitrary units)
*
*
¶
CREB
PKA
BetA
+++
++–
+––
CTL BetA siCREBBetA/CREB
CTL BetA siCREBBetA/CREB
CTL BetA siCREB
BetA/CREB
CTL BetA siCREBBetA/CREB
CTL BetA siCREBBetA/CREB
Fig. 4. BetA-mediated SOD2 suppression was due to direct inhibition of CREB phosphorylation. (A) HBV-infected hepatocytes were treated
with control (CTL), 15 lgÆmL
)1
BetA, BetA with CREB overexpression (BetA ⁄ CREB›) or siRNA for CREB (siCREB) for 48 h; the chromatin
from treated cells was immunoprecipitated with CREB antibody, and the SOD2 promoter that covers the CREB-binding motif was amplified
by quantitative PCR (qPCR). (B–E) The cells treated as above were used for measurement of CREB protein level (B), pCREB protein level
(C), SOD2 mRNA level (D), and SOD2 protein level (E). *P < 0.05 versus CTL for (A)–(E). (F) In vitro-purified proteins were phosphorylated
by PKA in the presence or absence of BetA, and pCREB was measured by western blotting. *P < 0.05 versus first panel; –P < 0.05 versus
second panel.
Betulinic acid inhibits hepatitis B virus D. Yao et al.
2604 FEBS Journal 276 (2009) 2599–2614 ª 2009 The Authors Journal compilation ª 2009 FEBS
decreased amount of phosphorylated CREB (or
decreased CREB activity) downregulates SOD2
expression through the CREB-binding motif on the

SOD2 promoter.
BetA suppresses HBx and translocates HBx to
mitochondria
We further examined the effect of BetA on HBx from
HBV-infected hepatocytes. The cells treated with con-
trol, BetA or siCREB were isolated into mitochon-
drial and cytosolic fractions for western blotting
analysis. As shown in Fig. 5A, the level of HBx pro-
tein was decreased in total lysates and cytosolic frac-
tions but increased in mitochondrial fractions after
BetA treatment, and siCREB mimicked the effect of
BetA, indicating that BetA treatment not only sup-
pressed HBx expression, but also translocated HBx
into mitochondria. We further measured cyto-
chrome c release for the treated cells. As shown in
Fig. 5B, the cytochrome c level was substantially
increased in cytosolic fractions after BetA or siCREB
treatment as compared with control, was decreased in
mitochondria, but was unchanged in total lysates.
This suggests that BetA-mediated cytochrome c release
and apoptosis may be associated with HBx transloca-
tion to mitochondria.
The BetA-mediated proapoptotic effect depends
on BetA-induced SOD2 suppression in
HBV-infected cells
In order to further determine the mechanisms of
BetA-induced HBx translocation and proapoptotic
effects, we measured the BetA-induced cytotoxicity in
different cells; as shown in Fig. 5C,D, BetA slightly
increased caspase-3 activity (Fig. 5C) and the apop-

tosis rate (Fig. 5D) in WT cells, and a similar effect
was observed in WT hepatocytes overexpressing
HBx; overexpression of CREB could not abolish this
effect, suggesting that the BetA-induced basal toxic
effect in WT cells is not due to BetA-induced SOD2
suppression, and HBx alone is not directly involved
in BetA-induced basal toxicity. On the other hand,
the BetA-induced toxicity was substantially increased
in HBV-infected hepatocytes as compared with WT
cells, and this effect was mostly abolished by overex-
pression of CREB, suggesting that, in HBV-infected
cells, BetA-induced toxicity is due to SOD suppres-
sion. We next measured SOD2 expression in different
0
300
A B C
D
E
F
200
100
Total lysate Mitochondria Cytosol
HBx protein level by Western blot
(Arbitrary units)
CTL
BetA
siCREB
*
*
*

*
*
*
0
300
200
100
CTL
BetA
WT WT-HBx HBV
Caspase-3 activity (pmol·min
–1
·mg
–1
)
*
*
*
*
*
¶
*
WT WT-HBx HBV
0
5
10
15
20
Apoptosis rate (%)
CTL

BetA
*
*
*
*
*
¶
*
0
300
200
100
Total lysate Mitochondria Cytosol
CTL
BetA
siCREB
CytC protein level by Western blot
(Arbitrary units)
*
*
*
*
HBV
CTL BetA
HBx protein in mitochondria by Western blot
(Arbitrary units)
*
?
0
400

300
200
100
WT WT-HBx HBV
CTL
BetA
0
400
300
200
100
SOD2 mRNA level by qPCR
*
*
(Arbitrary units)
Fig. 5. BetA reduces the level of HBx and translocates HBx to mitochondria through SOD2 suppression and subsequent ROS generation.
(A, B) HBV-infected hepatocytes were treated with 15 lgÆmL
)1
BetA for 48 h, the cells were separated as mitochondrial and cytosolic frac-
tions, and the protein levels were measured by western blotting. (A) HBx protein. (B) Cytochrome c (CytC) protein. *P < 0.05 versus control
(CTL) group. (C–E) WT hepatocytes, WT cells overexpressing HBX (WT-HBx cells) or HBV-infected hepatocytes were treated with either
CTL, 15 lgÆmL
)1
BetA or BetA with CREB overexpression (BetA ⁄ CREB›) for 48 h. (C) Intracellular caspase-3 activity. *P < 0.05 versus CTL;
–P < 0.05 versus BetA in the HBV-infected group. (D) Apoptosis rate determined by TUNEL assay. *P < 0.05 versus CTL; –P < 0.05 versus
BetA in the HBV-infected group. (E) SOD2 mRNA level. *P < 0.05 versus CTL in the WT group. (F) The mitochondrial fraction from the HBV-
infected hepatocytes treated as above was used for analysis of HBx by western blotting. The WT HBx group shows no detectable bands in
mitochondria (data not shown). *P < 0.05 versus CTL.
D. Yao et al. Betulinic acid inhibits hepatitis B virus
FEBS Journal 276 (2009) 2599–2614 ª 2009 The Authors Journal compilation ª 2009 FEBS 2605

cells; as shown in Fig. 5E, the basal level of SOD2
was not changed in WT cells or WT hepatocytes
overexpressing HBx, whereas the SOD2 level was
substantially increased in HBV-infected cells as com-
pared with WT cells; this increase was totally nor-
malized by BetA, and overexpression of CREB
minimized the effect of BetA, suggesting that BetA-
induced toxicity in HBV-infected cells is due to
BetA-mediated SOD suppression. Finally, we mea-
sured HBx translocation to mitochondria in different
cells, as shown in Fig. 5F. In WT hepatocytes over-
expressing HBx, HBx was not found in mitochondria
at all in the presence of BetA (data not shown),
whereas in HBV-infected cells, BetA-induced HBx
translocation was totally abolished by CREB overex-
pression, suggesting that BetA-induced SOD2
suppression and subsequent ROS generation is the
driving force for HBx transcloation to mitochondria.
The BetA-mediated inhibitory effect on HBV is
due to SOD2 suppression and subsequent ROS
generation
We previously found that BetA suppresses SOD2
expression by inhibiting CREB phosphorylation, with
subsequent ROS overgeneration. Here, we further
investigated the potential effect of SOD2 on HBV
replication. The HBV-infected hepatocytes were trea-
ted with BetA, or BetA with SOD2 overexpression,
or siRNA for SOD2 (siSOD2) alone, and the related
biomedical parameters were measured. As shown in
Fig. 6, the levels of SOD2 mRNA (Fig. 6A) and

SOD2 protein (Fig. 6B) decreased to 33% and 46%,
respectively, after BetA treatment, BetA treatment
with SOD2 overexpression caused no difference in
SOD2 level, and siSOD2 mimicked the effect of
BetA. We also measured ROS formation (Fig. 6C)
*
*
HBsAg level
(Arbitrary units)
HBeAg level
(Arbitrary units)
*
*
*
*
0
200
AE
F
G
H
B
C
D
100
0
200
100
0
200

100
0
200
100
0
200
100
SOD2 mRNA level by qPCR
(Arbitrary units)
0
200
100
300
200
100
SOD2 mRNA level by Western blot
(Arbitrary units)
*
?
*
0
ROS formation
(Arbitrary units)
*
*
0
30
20
10
Apoptosis rate

by TINEL assay (%)
*
*
CTL
BetA
si
SOD2
C
TL
BetA
si
SOD2
HBV DNA level by qPCR
(Arbitrary units)
*
*
HBx protein level
(Arbitrary units)
*
*
Fig. 6. BetA-mediated inhibitory effect on
HBV through SOD2 suppression and ROS
generation. HBV-infected hepatocytes were
treated with control (CTL), 15 lgÆmL
)1
BetA,
BetA with CREB overexpression (BetA ⁄ -
CREB›) or siCREB for 48 h, and the cells
were used for measurement of the indicated
parameters. (A) SOD2 mRNA level. (B) SOD2

protein level. (C) ROS generation. (D) Apopto-
sis rate determined by TUNEL assay. (E)
HBsAg secreted from cell culture medium.
(F) HBeAg secreted from cell culture med-
ium. (G) HBV DNA from treated cells was
measured by real-time quantitative PCR. (H)
HBx protein level was measured by western
blotting and quantitated. *P < 0.05 versus
the CTL group.
Betulinic acid inhibits hepatitis B virus D. Yao et al.
2606 FEBS Journal 276 (2009) 2599–2614 ª 2009 The Authors Journal compilation ª 2009 FEBS
and apoptosis (Fig. 6D); BetA treatment substantially
increased ROS generation and the apoptosis rate,
SOD2 overexpression in the presence of BetA mini-
mized the effect, and siSOD2 mimicked the effect of
BetA. We next measured the effect of these treat-
ments with different SOD2 expression levels on HBV
replication; the results showed that BetA alone sub-
stantially inhibited HBV replication, including hepati-
tis B surface antigen level (HBsAg) level (Fig. 6E),
hepatitis B external core antigen (HBeAg) level
(Fig. 6F), HBV DNA (Fig. 6G), and HBx protein
expression (Fig. 6H), whereas a combination of BetA
and SOD2 overexpression totally abolished the BetA-
mediated HBV-inhibitory effect. On the other hand,
SOD2 knockdown (siSOD2) mimicked the BetA-
induced inhibitory effect. This suggests that BetA-
induced ROS generation plays an important role in
HBV inhibition; scavenging of ROS by overexpres-
sion of the antioxidant enzyme SOD2 might not

be beneficial, but worsen the HBV infection, whereas
the increase in ROS generation caused by direct
SOD2 knockdown could achieve similar HBV-inhibi-
tory effects.
BetA mimics the PC-induced inhibitory effect on
HBV in mice
In order to verify that BetA or PC extract does not
alter general liver function and has no toxic effects
in healthy liver, nontransgenic mice were employed
to evaluate the proapoptotic effect. Thirty male non-
HBV-transgenic mice were randomly separated into
three groups (10 in each). Experimental groups
received either purified BetA (2 mgÆkg
)1
)orPC
crude extracts (50 mgÆkg
)1
), whereas the control
group received only vehicle. Drugs or vehicle were
added to the normal food and mixed for feeding.
After 3 months, mice were killed by decapitation.
The liver tissues were collected for measurement of
biomedical parameters: (a) SOD2 mRNA level; (b)
enzymatic activities of caspase-3 and SOD2; (c)
superoxide release; and (d) enzymatic activities of
alanine aminotranferase and aspartate aminotrans-
ferase. We found that neither BetA nor PC extract
had significant cytotoxic effects on hepatocytes from
mice (data not shown). In addition, we have previ-
ously found that BetA isolated from PC inhibits

HBV replication in vitro by SOD2 suppression, which
is similar to the effect that PC had in hepatitis B
patients in our preliminary observation (data not
shown). Here, we used HBV-transgenic mice to
determine whether BetA could achieve the same
inhibitory effect. As shown in Fig. 7A, both BetA
and PC significantly reduced HBsAg and HBeAg
serum levels and HBV DNA replication. Also, both
BetA and PC substantially decreased SOD2 mRNA
expression, whereas CREB mRNA showed no
changes (Fig. 7B). In addition, protein levels of
SOD2 and pCREB were substantially reduced after
BetA and PC treatment, whereas no changes were
found in CREB total protein level (Fig. 7C). We
also examined the enzymatic activities, and showed
that both BetA and PC not only decreased SOD2
activity, but also increased caspase-3 activity, indicat-
ing increased cytotoxicity with apoptosis rate
(Fig. 6D). Finally, we examined the levels of super-
oxide release in different tissues (Fig. 6E,F); both
BetA and PC specifically increased superoxide anion
generation in liver tissue, but had little effect in
aorta, and no effect at all in kidney and brain, indi-
cating that both BetA-mediated and PC-mediated
HBV inhibition are due to specifically decreased
SOD2 expression with subsequent ROS generation in
liver tissue.
Discussion
This study demonstrates that BetA inhibits HBV repli-
cation by suppression of SOD2 expression with subse-

quent mitochondrial ROS overgeneration, with
promising HBV clearance in both in vitro and in vivo
mouse experiments. This is the first time that we have
shown the potential effects and possible mechanism of
HBV inhibition by BetA.
BetA-mediated selective cytotoxicity in
HBV-infected hepatocytes
We have found that BetA has little cytotoxic effect on
WT hepatocytes, but shows a selective cytotoxic effect
on HBV-infected hepatocytes. In addition, our data
showed that the basal level of SOD2 was not changed
in WT or WT hepatocytes overexpressing HBx,
whereas the SOD2 level was substantially increased in
HBV-infected cells as compared with WT cells, that
this increase was totally normalized by BetA, and that
overexpression of CREB could minimize the effect of
BetA (Fig. 5E), suggesting that BetA-induced toxicity
in HBV-infected cells is due to BetA-mediated SOD
suppression. This indicates that HBV infection in
HBV-infected cells specifically increases SOD2 expres-
sion, even though the detailed mechanisms are still
unknown. Furthermore, the basal SOD2 protein level
in HBV-infected cells is much higher than in WT cells,
and it is reasonable that the HBV-infected cells with
high levels of SOD2 expression should be more
D. Yao et al. Betulinic acid inhibits hepatitis B virus
FEBS Journal 276 (2009) 2599–2614 ª 2009 The Authors Journal compilation ª 2009 FEBS 2607
susceptible to BetA-mediated SOD2 suppression, and,
subsequently, the SOD2 suppression-mediated large
increase in mitochondrial ROS generation may further

induce mitochondrial dysfunction and apoptosis [24].
Given the fact that HBV-infected cells are more sus-
ceptible to BetA-induced SOD2 suppression, and
BetA-induced SOD2 suppression could directly inhibit
HBV replication, as shown in Fig. 6, we conclude that
BetA could be a good candidate for anti-HBV drug
development.
BetA-mediated CREB dephosphorylation
As BetA could cause CREB dephosphorylation at
Ser133 both in vivo and in vitro, and a mutated form
of CREB with an Ala substitution for Ser133 has been
reported to be a negative transcriptional regulator,
BetA-induced dephosphorylation of CREB could act
as a repressor of SOD2 gene transcription directly [25].
CREB, as a direct substrate of both PKA [21,26] and
protein kinase D (PKD) [27], could be phosphorylated
0
100
200
Protein level by Western blot
(Arbitrary units)
CLT BAte PC
*
*
*
*
0
50
100
150

200
Superoxide anions release (n
M
)
CTL BetA PC
*
*
*
*
CTL BetA
PC
*
*
0
50
100
150
0
50
100
150
HBsAg HBeAg HBV-DNA
HBV expression (Arbitrary units)
mRNA expression (Arbitrary units)
CTL BetA PC
*
*
*
*
*

*
0
100
200
300
400
Enzyme activity (Arbitrary units)
CTL
BetA
PC
*
*
*
*
Caspase-3 SOD2
Liver Aorta Kidney Brain
CTL BetA PC
CREB SOD2
CREB
pCREB
SOD2
16.00 µm
16.00 µm
16.00 µm
AB
C
D
E
F
Fig. 7. BetA-mediated HBV inhibitory effect in mice through SOD2 suppression and ROS generation. HBV-transgenic mice were treated

with either vehicle [control (CTL)], BetA or PC crude extracts for 3 months, the mice were killed, and the medical parameters from blood or
different tissues were measured. (A) HBsAg, HBeAg and HBV DNA were measured from blood. (B) mRNA expression for CREB and SOD2
was measured by quantitative PCR from liver tissue. (C) Protein levels for CREB, pCREB and SOD2 were measured by western blotting
from liver tissue and quantitated. (D) Enzyme activities for caspase-3 and SOD2 from liver tissue were measured and expressed as arbitrary
units. (E) Superoxide anion release from different tissues was measured. (F) Representative images for in vivo superoxide staining from liver
tissue. *P < 0.05 versus CTL group.
Betulinic acid inhibits hepatitis B virus D. Yao et al.
2608 FEBS Journal 276 (2009) 2599–2614 ª 2009 The Authors Journal compilation ª 2009 FEBS
with the recruitment of coactivator CBP ⁄ p300 [21,28];
our results showed that PKA-induced in vitro CREB
phosphorylation was directly inhibited by BetA
(Fig. 4F), whereas no effect on PKD-induced phos-
phorylation was observed (data not shown), and the
expression of neither PKA nor PKD was changed after
BetA treatment (data not shown), suggesting that
BetA-induced CREB dephosphorylation is due to
direct inhibition of PKA by BetA.
BetA-induced HBx suppression and translocation
HBx protein, which is encoded by the X gene of the
HBV genome, could stimulate many different signal
transduction pathways, or directly interact with tran-
scription factors, resulting in several different biologi-
cal activities in HBV-associated liver disease [29,30].
Also, HBx could alter mitochondrial function through
direct association with mitochondria [31,32]. In BetA-
treated HBV-infected hepatocytes, we found that the
total HBx protein level was substantially decreased,
indicating that BetA may inhibit HBV-associated
hepatocarcinogenesis through suppression of HBx
expression; this may partly explain its potential anti-

tumor effect. Furthermore, this is the first time that
HBx has been found to be translocated into mitochon-
dria from the cytosol in BetA-treated HBV-infected
hepatocytes; given the fact that HBx could induce
apoptosis [33,34] and alter mitochondrial function by
inhibiting the mitochondrial electron transport chain
and oxidative phosphorylation (complexes I, III, IV,
and V) [31], we suppose that BetA-mediated HBx sup-
pression and translocation worsens the mitochondrial
dysfunction, which may further trigger apoptosis. As
SOD2 expression could totally abolish the BetA-
induced HBx suppression (Fig. 5A), we suppose that
HBx translocation to mitochondria could be associated
with BetA-induced SOD2 suppression. In order to
determine the possible role and effect of HBx in this
procedure, as shown in Fig. 5C–F, HBx protein was
overexpressed in WT hepatocytes overexpressing HBx,
and the related proapoptotic effect was measured. We
found that HBx alone (instead of full HBV infection)
showed a similar effect in WT cells after BetA treat-
ment, and caused small increases in caspase-3 activity
(Fig. 5C) and apoptosis rate (Fig. 5D), no increase in
SOD2 expression (Fig. 5E), and no HBx translocation
into mitochondria (data not shown). On the other
hand, in HBV-infected hepatocytes, SOD2 expression
was substantially increased, and became sensitive to
the BetA-induced proapoptotic effect, the HBx was
translocated into mitochondria, and the effect was nor-
malized by CREB overexpression (Fig. 5F), suggesting
that HBx alone does not directly contribute to BetA-

induced SOD2 suppression and HBV inhibition,
whereas HBx translocation to mitochondria could be a
consequence of BetA-induced SOD2 suppression and
subsequent ROS generation.
BetA-mediated mitochondrial redox balance
We found that BetA has a strong redox effect through
suppression of SOD2 expression with subsequent mito-
chondrial ROS generation; even though ROS genera-
tion could cause cell damage, our results show that
BetA-mediated ROS generation is favorable for HBV
clearance, as SOD2 overexpression could abolish
BetA-induced HBV inhibition, whereas siSOD2 mim-
icked BetA-induced HBV inhibition, suggesting that
the BetA-mediated redox effect could be both benefi-
cial and detrimental to cells; the most important thing
is to maintain a balance that achieves the best redox
environment for living organisms [24,35].
BetA-mediated selective superoxide release in
liver tissue
An interesting result was found from the mouse experi-
ments with specifically increased superoxide anion
release in liver tissue under BetA treatment, instead of
in other tissues. As superoxide formation may directly
correspond to SOD2 expression, we measured SOD2
protein levels in some other tissues, and found that
liver tissue has much higher basal SOD2 levels than
other tissues, including aorta, kidney, heart, and lung,
with the exception of a higher SOD2 level in brain
(data not shown). Given the fact that the liver accepts
most of the circulating BetA under dietary conditions

for subsequent deposition and detoxification as the
first target organ for xenobiotics, the high susceptibi-
lity of the liver could due to its higher SOD2 basal
expression and higher accessibility to BetA. On the
other hand, in brain tissue, the circulating BetA can
hardly gain access, owing to the protection of the
blood–brain barrier, so the brain is insensitive to BetA
treatment, even with higher SOD2 expression levels.
Taken together, these findings demonstrate that
BetA could achieve an impressive HBV-inhibitory
effect by specific suppression of SOD2 expression and
modulation of the mitochondrial redox balance. The
BetA used in this work was isolated and purified
from PC, a traditional Chinese herb that has been
succesfully used for the treatment of hepatitis B
patients, according to a secret recipe. According to the
Chinese traditional medicine theory: ‘Everything has
its own enemy from nature,’ which essentially means
D. Yao et al. Betulinic acid inhibits hepatitis B virus
FEBS Journal 276 (2009) 2599–2614 ª 2009 The Authors Journal compilation ª 2009 FEBS 2609
that everything within the world not only has its own
method of survival, but also has its own method of
destruction, thus preserving the balance of nature. In
fact, we can derive elements for good health from nat-
ure if we utilize it properly. Given the fact that PC is
well known and relatively nontoxic, with the promising
effects presented by its component BetA, PC scould be
a good candidate for anti-HBV drug development.
Experimental procedures
Materials and methods

HBV-transgenic mice (male, 6–8 weeks old, 18–22 g) were
obtained from the Transgenic Engineering Laboratory in
the Infectious Disease Center (Guangzhou, China) [36,37].
The mouse genotypes were identified from both positive
HBsAg and HBeAg; the serum HBV virus load in these
mice was 10
4
–10
5
copies per mL; HBV DNA, HBV mRNA
and hepatitis B core antigen in hepatocytes were shown to
be positive by Southern blotting, northern blotting, and
immunohistochemistry, respectively. Animals showed no
evident cytopathological changes in the liver, with normal
serum alanine aminotransferase levels. Animals were
housed in a pathogen-free room under strict barrier condi-
tions. All procedures used in dealing with the experimental
animals were approved by our institutional animal care and
use committee.
Primary hepatocytes from WT or HBV-transgenic mice
were prepared from residual liver tissue from related mice by
enzymatic dissociation; the isolation protocol was approved
by the Ethics Committee, Tongji Medical School. Cells were
kept in William’s E medium supplemented with 5% fetal
bovine serum and antibiotics at 37 °Cin5%CO
2
[38].
HBx DNA amplified by PCR from HBV genomic
DNA, CREB-1 cDNA and SOD2 cDNA were subcloned
into pcDNA3.1 for overexpression of HBx, CREB-1, and

SOD2, respectively. The CREB-1 cDNA was further sub-
cloned into the pGEX-4T vector (no. 27-4580-01; GE
Healthcare, Shanghai, China) for expression of CREB
protein in Escherichia coli BL21 cells, and this was fol-
lowed by glutathione S-transferase protein purification
with the MagneGST Protein Purification System
(no. V8600; Promega, Beijing, China) and thrombin prote-
ase digestion. The SOD2 promoter was amplified from
human genomic DNA and subcloned into the pGL3-basic
vector (no. E1751; Promega) to construct the SOD2
reporter plasmid. For mapping of the SOD2 promoter
response element, the related deletion or point mutation
constructs were generated by PCR methods or a site-
directed mutagenesis kit (no. Q9280; Promega). Detailed
information on those plasmids is available upon request.
Antibody against HBx (no. RD981038100) was obtained
from BioVendor Laboratories Ltd (Guangzhou, China),
and rabbit voltage-dependent anion selective channel
1 ⁄ porin antibody (V2139) was obtained from Sigma.
Antibodies against b-actin (sc-47778), CREB-1 (sc-58),
pCREB-1 (Ser133 phosphorylation, sc-7978) and SOD2
(sc-30080) were obtained from Santa Cruz (Beijing, China).
The mitochondrial and cytosolic fractions were pre-
pared and characterized using differential centrifugation
as previously described [39]. Protein concentrations were
measured with the BioRad protein assay kit (Bradford
method), according to the manufacturer’s instructions.
Small interfering RNAs for CREB-1 (no. s3491), SOD2
(no. 9052) and negative control (no. AM4636) were
obtained from Ambion (Shanghai, China). The SOD2

activity was measured with the SOD Detection Kit
(no. CSOD100-2; Cell Technology Inc., Shanghai, China).
The plasmid DNA and siRNA were transfected
with Lipofectamine Reagent (Invitrogen, Beijing, China).
Luciferase activity assays were carried out using the
Dual-Luciferase Assay System (Promega), and transfec-
tion efficiencies were normalized using a cotransfected
Renilla plasmid.
BetA isolation and purification
BetA was isolated from PC collected from a special area of
China (Yichang, Hubei), in order to achieve reliable and
repeatable clinical effects. The root of PC was washed,
dried at room temperature (avoiding direct sunlight), and
cut into small pieces. Five hundred grams of PC was then
added to 495 mL of distilled water and 5 mL of fish oil
(Sigma), and incubated at 37 °C for no less than 4 h, with
continuous stirring. The temperature was then increased
slowly up to boiling point, and maintained for 10 min for
decoction. The solution was sterile filtered, concentrated
and lyophilized into powder for PC crude extracts; the
powder was mixed with a standard powdered rodent chow
for animal treatment, or the PC crude extracts were further
dissolved in 3 : 1 acetonitrile ⁄ methanol, purified, and char-
acterized with an HPLC system coupled to a negative IES
mass spectrometer [40].
Cell viability and MTT assay
Cells were pooled in 12-well plates, following exposure to
different treatments as indicated, at 80% confluence. Cell
viability was analyzed by the MTT reduction assay.
Briefly, cells in each well were aspirated and washed with

NaCl ⁄ P
i
, and 0.2 mL of 0.3 mgÆmL
)1
MTT solution was
then added at 25 °C for 3 h. Thereafter, the precipitated
blue formazan product was extracted by incubating sam-
ples with 0.1 mL of 10% SDS (dissolved in 0.01 m HCl)
overnight at 37 °C. The absorbances of formazan concen-
trations were determined at 570 nm, then normalized by
cell numbers and expressed as A ⁄ 10
6
cells.
Betulinic acid inhibits hepatitis B virus D. Yao et al.
2610 FEBS Journal 276 (2009) 2599–2614 ª 2009 The Authors Journal compilation ª 2009 FEBS
DNA synthesis evaluated by [
3
H]thymidine
incorporation
Cell proliferation was evaluated as the rate of DNA synthe-
sis by [
3
H]methylthymidine incorporation [41]. Cells were
pooled in 24-well plates up to 80% confluence, and differ-
ent concentrations of BetA were then added and incubated
for 48 h. At the end of the treatment, cells were incubated
with serum-free medium containing [
3
H]methylthymidine
(0.5 lCi per well) for 2 h, and then washed twice with

NaCl ⁄ P
i
. Cellular DNA was precipitated with 10% trichlo-
roacetic acid and solubilized with 0.4 m NaOH (0.5 mL per
well). Incorporation of [
3
H]methylthymidine into DNA was
measured in a scintillation counter and was normalized to
protein concentration.
Reverse transcription reaction and real-time PCR
Total RNA from treated cells was extracted with an
RNeasy Mini Kit (Qiagen, Shanghai, China), and the RNA
was reverse transcribed with an Omniscript RT kit (Qia-
gen). Real-time quantitative PCR was performed on an
iCycler iQ (Bio-Rad, Shanghai, China) with the Quantitect
SYBR Green PCR kit (Qiagen). Two-step PCR was per-
formed at 95 °C for 5 min, followed by 45 cycles of dena-
turation at 95 °C for 30 s, and annealing and extension at
60 °C for 60 s. The fluorescence was detected at the end of
each extension step. The results are normalized to b-actin.
Western blotting
Cells were lysed in ice-cold lysis buffer (0.137 m NaCl, 2 mm
EDTA, 10% glycerol, 1% NP-40, 20 mm Tris base, pH 8.0)
plus protease inhibitor cocktail (Sigma). The proteins were
separated by 10% SDS ⁄ PAGE and transferred to a poly
(vinylidene difluoride) membrane. The membrane was then
incubated with appropriate antibodies, washed, and incu-
bated with horseradish peroxidase-labeled secondary antibod-
ies. The blots were visualized with the ECL+plus Western
Blotting Detection System (Amersham, Shanghai, China),

and quantitated by imagequant. The final results were
normalized to b-actin or porin (for mitochondrial protein).
Chromatin immunoprecipitation
Treated cells were crosslinked using 1% formaldehyde for
20 min and terminating with 0.1 m glycine. Cell lysates
were sonicated and centrifuged at 18 000 g for 12 min to
get supernatant. Five hundred micrograms of protein was
precleared by BSA ⁄ salmon sperm DNA plus preimmune
IgG and a slurry of protein A–agarose beads, as previously
described [42]. Immunoprecipitations were performed with
the indicated antibodies, BSA ⁄ salmon sperm DNA, and a
50% slurry of protein A–agarose beads. The immunopre-
cipitates were washed and eluted, and then incubated with
0.2 mgÆmL
)1
proteinase K for 2 h at 42 °C, and then for
6 h at 65 °C, to reverse the formaldehyde crosslinking.
DNA fragments were recovered by phenol ⁄ chloroform
extraction and ethanol precipitation. A 150 bp fragment
from the human SOD2 promoter was amplified by real-time
quantitative PCR.
In vitro phosphorylation of CREB at Ser133
Purified CREB was phosphorylated by the catalytic subunit
of PKA (no. P2645; Sigma) by incubating 2.0 lm CREB in
a reaction mixture containing 4.0 lm ATP (or UTP for
mock reactions), 8 mm MgCl
2
and 100 U of PKA in
25 mm NaCl ⁄ P
i

(pH 6.6) at 30 °C for 2 h in the presence
or absence of 15 lgÆmL
)1
BetA. The PKA was heat-inacti-
vated at 75 °C for 10 min. The level of pCREB was deter-
mined by western blotting [43].
Measurement of mitochondrial function
The intracellular ATP level was determined by the lucif-
erin ⁄ luciferase-induced bioluminescence system. An ATP
standard curve was generated at concentrations of 10
)12
–
10
)3
m. Intracellular ATP levels were calculated and
expressed as nmol ⁄ mg protein. MMP (DW
m
) was measured
by the 3,3¢-dihexiloxadicarbocyanine (DiOC
6
) method. Cells
were trypsinized, resuspended, and incubated with 0.2 lm
DiOC
6
for 20 min at 37 °C, and cells were then treated
with 1 lm propidium iodide for 30 min. The DiOC
6
fluo-
rescence was measured by fluorescence activated cell sorting
at an excitation ⁄ emission wavelength of 485 ⁄ 500 nm [39].

Evaluation of apoptosis
Apoptosis was evaluated by TUNEL assay using an In Situ
Cell Death Detection Kit (Roche, Shanghai, China). Cells
were fixed in 4% paraformaldehyde, and labeled with
TUNEL reagents. Stained cells were photographed with a
fluorescence microscope, and further quantified by fluores-
cence activated cell sorting analysis. Caspase-3 activity was
determined with an ApoAlert caspase assay kit (Clontech,
Beijing, China). Treated cells were harvested, and 50 lgof
protein was incubated with the fluorogenic peptide substrate
Ac-DEVD-7-amino-4-trifluoromethyl coumarin. The initial
rate of free Ac-DEVD-7-amino-4-trifluoromethyl coumarin
release was measured using an FLx800 microplate reader
(Bio-Tek, Guangzhou, China) at excitation ⁄ emission wave-
lengths of 380 ⁄ 505 nm, and enzyme activity was calculated
as pmol ⁄ min ⁄ mg or as arbitrary units [39].
Measurement of oxidative stress
Intracellular ROS generation was determined by using oxi-
dation of 2¢,7¢-dihydrochlorofluorescein-diacetate. Treated
D. Yao et al. Betulinic acid inhibits hepatitis B virus
FEBS Journal 276 (2009) 2599–2614 ª 2009 The Authors Journal compilation ª 2009 FEBS 2611
cells were washed and incubated with 0.1 mL of 10 lm
2¢,7¢-dihydrochlorofluorescein-diacetate, and the fluorescence
was measured at excitation ⁄ emission wavelengths of 485 ⁄
530 nm, using an FLx800 microplate fluorescence reader
(Bio-Tek). The data were normalized to arbitrary units [39].
Measurement of HBV replication
HBV-infected hepatocytes were seeded in 24-well plates,
and then subjected to BetA treatment for 48 h. Cell num-
bers were determined by Trypan blue exclusion. Secretion

of HBsAg and HBeAg in cultured medium or serum
from mice was assayed by using commercially available kits
(Abbott, Shanghai, China) as previously described [44]. The
DNA from cells or serum samples was extracted with phe-
nol ⁄ chloroform ⁄ isoamyl alcohol (25 : 24 : 1), and precipi-
tated with ethanol after digestion at 37 °C for 2 h with
proteinase K (0.5 mgÆmL
)1
) in the presence of SDS (0.5%).
The HBV DNA level was measured by real time-PCR,
using human b -actin as control [45].
Effect of BetA on HBV-transgenic mice
Sixty male HBV-transgenic mice were randomly separated
into three groups (20 in each). Experimental groups
received either purified BetA (2 mgÆkg
)1
) or PC crude
extracts (50 mgÆkg
)1
), whereas the control group received
only vehicle. Drugs or vehicle were added to the normal
food and mixed for feeding to the experimental mice.
After 3 months, mice were killed by decapitation. Blood
(with EDTA as anticoagulant) and other tissues, including
brain, heart, aorta, and kidney, were collected for serum
HBV replications or other biomedical parameters as indi-
cated. Serum HBsAg, HBeAg and HBV DNA were ana-
lyzed as described above, gene expression was measured
by quantitative PCR and western blotting, and superoxide
release from tissues was measured as previously described

[39]. The in vitro staining of superoxide anions (O
2
·
)
) was
performed with the oxidative fluorescent dye dihydroethi-
dium [46,47]. Briefly, fresh and unfixed liver tissues were
frozen and cut in a cryostat into 30 lm sections and
placed on glass slides. Samples were then incubated at
room temperature for 30 min with dihydroethidium
(0.002 mmolÆL
)1
) and protected from light. Images were
obtained using a laser (krypton ⁄ argon) scanning confocal
microscope with fluorescence excitation ⁄ emission at
488 ⁄ 610 nm.
Statistical analysis
Data are given as mean ± standard deviation. All experi-
ments were performed at least three times. All analyses
were performed using spss 15.0 statistical software. Stu-
dent’s unpaired t-test or ANOVA were used to determine
the statistical significance of different groups. A P-value
< 0.05 was considered to be significant.
Acknowledgements
This work was supported by the National Natural Sci-
ence Foundation of China (Project No. 30270563) and
the Science Research Foundation of Health Depart-
ment of Hubei Province (Project No. JX2A04).
References
1 Lau JY & Wright TL (1993) Molecular virology and

pathogenesis of hepatitis B. Lancet 342, 1335–1340.
2 Modi AA, Wright EC & Seeff LB (2007) Complemen-
tary and alternative medicine (CAM) for the treatment
of chronic hepatitis B and C: a review. Antivir Ther 12,
285–295.
3 Liu WK, Ho JC, Cheung FW, Liu BP, Ye WC & Che
CT (2004) Apoptotic activity of betulinic acid deriva-
tives on murine melanoma B16 cell line. Eur J Pharma-
col 498, 71–78.
4 Bi Y, Xu J, Wu X, Ye W, Yuan S & Zhang L (2007)
Synthesis and cytotoxic activity of 17-carboxylic acid
modified 23-hydroxy betulinic acid ester derivatives.
Bioorg Med Chem Lett 17 , 1475–1478.
5 Ye WC, Ji NN, Zhao SX, Liu JH, Ye T, McKervey
MA & Stevenson P (1996) Triterpenoids from Pulsatilla
chinensis. Phytochemistry 42, 799–802.
6 Pisha E, Chai H, Lee IS, Chagwedera TE, Farnsworth
NR, Cordell GA, Beecher CW, Fong HH, Kinghorn
AD, Brown DM et al. (1995) Discovery of betulinic
acid as a selective inhibitor of human melanoma that
functions by induction of apoptosis. Nat Med 1, 1046–
1051.
7 Kasperczyk H, La Ferla-Bruhl K, Westhoff MA,
Behrend L, Zwacka RM, Debatin KM & Fulda S
(2005) Betulinic acid as new activator of NF-kappaB:
molecular mechanisms and implications for cancer
therapy. Oncogene 24, 6945–6956.
8 Evers M, Poujade C, Soler F, Ribeill Y, James C, Lelie-
vre Y, Gueguen JC, Reisdorf D, Morize I, Pauwels R
et al. (1996) Betulinic acid derivatives: a new class of

human immunodeficiency virus type 1 specific inhibitors
with a new mode of action. J Med Chem 39, 1056–1068.
9 Yogeeswari P & Sriram D (2005) Betulinic acid and its
derivatives: a review on their biological properties. Curr
Med Chem 12, 657–666.
10 Fulda S, Scaffidi C, Susin SA, Krammer PH, Kroemer
G, Peter ME & Debatin KM (1998) Activation of mito-
chondria and release of mitochondrial apoptogenic fac-
tors by betulinic acid. J Biol Chem 273, 33942–33948.
11 Yen HC, Oberley TD, Vichitbandha S, Ho YS & St
Clair DK (1996) The protective role of manganese
Betulinic acid inhibits hepatitis B virus D. Yao et al.
2612 FEBS Journal 276 (2009) 2599–2614 ª 2009 The Authors Journal compilation ª 2009 FEBS
superoxide dismutase against adriamycin-induced acute
cardiac toxicity in transgenic mice. J Clin Invest 98,
1253–1260.
12 Lam EW, Zwacka R, Seftor EA, Nieva DR, Davidson
BL, Engelhardt JF, Hendrix MJ & Oberley LW (1999)
Effects of antioxidant enzyme overexpression on the
invasive phenotype of hamster cheek pouch carcinoma
cells. Free Radic Biol Med 27, 572–579.
13 Xu Y, Fang F, Dhar SK, St Clair WH, Kasarskis EJ &
St Clair DK (2007) The role of a single-stranded nucle-
otide loop in transcriptional regulation of the human
sod2 gene. J Biol Chem 282, 15981–15994.
14 Xu Y, Krishnan A, Wan XS, Majima H, Yeh CC,
Ludewig G, Kasarskis EJ & St Clair DK (1999)
Mutations in the promoter reveal a cause for the
reduced expression of the human manganese super-
oxide dismutase gene in cancer cells. Oncogene 18,

93–102.
15 Hurt EM, Thomas SB, Peng B & Farrar WL (2007)
Integrated molecular profiling of SOD2 expression in
multiple myeloma. Blood 109, 3953–3962.
16 Huang Y, Peng J, Oberley LW & Domann FE (1997)
Transcriptional inhibition of manganese superoxide
dismutase (SOD2) gene expression by DNA methyla-
tion of the 5¢ CpG island. Free Radic Biol Med 23, 314–
320.
17 Bedogni B, Pani G, Colavitti R, Riccio A, Borrello S,
Murphy M, Smith R, Eboli ML & Galeotti T (2003)
Redox regulation of cAMP-responsive element-binding
protein and induction of manganous superoxide dismu-
tase in nerve growth factor-dependent cell survival.
J Biol Chem 278, 16510–16519.
18 Galeotti T, Pani G, Capone C, Bedogni B, Borrello S,
Mancuso C & Eboli ML (2005) Protective role of
MnSOD and redox regulation of neuronal cell survival.
Biomed Pharmacother 59, 197–203.
19 Gonzalez GA & Montminy MR (1989) Cyclic AMP
stimulates somatostatin gene transcription by phosphor-
ylation of CREB at serine 133. Cell 59, 675–680.
20 Goodman RH & Smolik S (2000) CBP ⁄ p300 in cell
growth, transformation, and development. Genes Dev
14, 1553–1577.
21 Zhang X, Odom DT, Koo SH, Conkright MD,
Canettieri G, Best J, Chen H, Jenner R, Herbolsheimer
E, Jacobsen E et al. (2005) Genome-wide analysis of
cAMP-response element binding protein occupancy,
phosphorylation, and target gene activation in human

tissues. Proc Natl Acad Sci USA 102, 4459–4464.
22 Vallejo M, Gosse ME, Beckman W & Habener JF
(1995) Impaired cyclic AMP-dependent phosphorylation
renders CREB a repressor of C ⁄ EBP-induced transcrip-
tion of the somatostatin gene in an insulinoma cell line.
Mol Cell Biol 15, 415–424.
23 Bito H, Deisseroth K & Tsien RW (1996) CREB phos-
phorylation and dephosphorylation: a Ca(2+)- and
stimulus duration-dependent switch for hippocampal
gene expression. Cell 87, 1203–1214.
24 Costantini P, Jacotot E, Decaudin D & Kroemer G
(2000) Mitochondrion as a novel target of
anticancer chemotherapy. J Natl Cancer Inst 92,
1042–1053.
25 Lamph WW, Dwarki VJ, Ofir R, Montminy M &
Verma IM (1990) Negative and positive regulation by
transcription factor cAMP response element-binding
protein is modulated by phosphorylation. Proc Natl
Acad Sci USA 87, 4320–4324.
26 Sharma N, Lopez DI & Nyborg JK (2007) DNA bind-
ing and phosphorylation induce conformational altera-
tions in the kinase-inducible domain of CREB.
Implications for the mechanism of transcription func-
tion. J Biol Chem 282, 19872–19883.
27 Johannessen M, Delghandi MP, Rykx A, Dragset M,
Vandenheede JR, Van Lint J & Moens U (2007) Protein
kinase D induces transcription through direct phosphor-
ylation of the cAMP-response element-binding protein.
J Biol Chem 282, 14777–14787.
28 Chrivia JC, Kwok RP, Lamb N, Hagiwara M, Montmi-

ny MR & Goodman RH (1993) Phosphorylated CREB
binds specifically to the nuclear protein CBP. Nature
365, 855–859.
29 Waris G, Huh KW & Siddiqui A (2001) Mitochondri-
ally associated hepatitis B virus X protein constitu-
tively activates transcription factors STAT-3 and
NF-kappa B via oxidative stress. Mol Cell Biol 21,
7721–7730.
30 Bouchard MJ, Wang L & Schneider RJ (2006) Activa-
tion of focal adhesion kinase by hepatitis B virus HBx
protein: multiple functions in viral replication. J Virol
80, 4406–4414.
31 Lee YI, Hwang JM, Im JH, Lee YI, Kim NS, Kim
DG, Yu DY, Moon HB & Park SK (2004) Human hep-
atitis B virus-X protein alters mitochondrial function
and physiology in human liver cells. J Biol Chem 279,
15460–15471.
32 Huh KW & Siddiqui A (2002) Characterization of the
mitochondrial association of hepatitis B virus X pro-
tein, HBx. Mitochondrion 1, 349–359.
33 Miao J, Chen GG, Chun SY & Lai PP (2006) Hepati-
tis B virus X protein induces apoptosis in hepatoma
cells through inhibiting Bcl-xL expression. Cancer Lett
236, 115–124.
34 Tang H, Oishi N, Kaneko S & Murakami S (2006)
Molecular functions and biological roles of hepatitis B
virus x protein. Cancer Sci 97, 977–983.
35 Boekema EJ & Braun HP (2007) Supramolecular struc-
ture of the mitochondrial oxidative phosphorylation
system. J Biol Chem 282, 1–4.

36 Guidotti LG, Matzke B, Schaller H & Chisari FV
(1995) High-level hepatitis B virus replication in trans-
genic mice. J Virol 69 , 6158–6169.
D. Yao et al. Betulinic acid inhibits hepatitis B virus
FEBS Journal 276 (2009) 2599–2614 ª 2009 The Authors Journal compilation ª 2009 FEBS 2613
37 Chisari FV (1995) Hepatitis B virus transgenic mice:
insights into the virus and the disease. Hepatology 22,
1316–1325.
38 Patel RD, Hollingshead BD, Omiecinski CJ & Perdew
GH (2007) Aryl-hydrocarbon receptor activation regu-
lates constitutive androstane receptor levels in murine
and human liver. Hepatology 46, 209–218.
39 Yao D, Shi W, Gou Y, Zhou X, Yee Aw T, Zhou Y &
Liu Z (2005) Fatty acid-mediated intracellular iron
translocation: a synergistic mechanism of oxidative
injury. Free Radic Biol Med 39, 1385–1398.
40 Cheng X, Shin YG, Levine BS, Smith AC, Tomaszew-
ski JE & van Breemen RB (2003) Quantitative analysis
of betulinic acid in mouse, rat and dog plasma using
electrospray liquid chromatography ⁄ mass spectrometry.
Rapid Commun Mass Spectrom 17, 2089–2092.
41 Somasundaram K & El-Deiry WS (1997) Inhibition of
p53-mediated transactivation and cell cycle arrest by
E1A through its p300 ⁄ CBP-interacting region. Oncogene
14, 1047–1057.
42 Metivier R, Penot G, Hubner MR, Reid G, Brand H,
Kos M & Gannon F (2003) Estrogen receptor-alpha
directs ordered, cyclical, and combinatorial recruitment
of cofactors on a natural target promoter. Cell 115,
751–763.

43 Giebler HA, Loring JE, van Orden K, Colgin MA,
Garrus JE, Escudero KW, Brauweiler A & Nyborg JK
(1997) Anchoring of CREB binding protein to the
human T-cell leukemia virus type 1 promoter: a molecu-
lar mechanism of Tax transactivation. Mol Cell Biol 17,
5156–5164.
44 Uprichard SL, Boyd B, Althage A & Chisari FV (2005)
Clearance of hepatitis B virus from the liver of trans-
genic mice by short hairpin RNAs. Proc Natl Acad Sci
USA 102, 773–778.
45 Takehara T, Suzuki T, Ohkawa K, Hosui A, Jinushi
M, Miyagi T, Tatsumi T, Kanazawa Y & Hayashi N
(2006) Viral covalently closed circular DNA in a non-
transgenic mouse model for chronic hepatitis B virus
replication. J Hepatol 44, 267–274.
46 Hathaway CA, Heistad DD, Piegors DJ and Miller FJ
Jr (2002) Regression of atherosclerosis in monkeys
reduces vascular superoxide levels. Circ Res 90, 277–
283.
47 Miller FJ Jr, Gutterman DD, Rios CD, Heistad DD &
Davidson BL (1998) Superoxide production in vascular
smooth muscle contributes to oxidative stress and
impaired relaxation in atherosclerosis. Circ Res 82,
1298–1305.
Betulinic acid inhibits hepatitis B virus D. Yao et al.
2614 FEBS Journal 276 (2009) 2599–2614 ª 2009 The Authors Journal compilation ª 2009 FEBS