Synthesis of [3β-acetoxy-urs-12-en-28-oyl]-1-monoglyceride and investigation on its anti tumor effects against BGC-823

Article abstract of DOI:10.1016/j.bmc.2011.05.017

Ursolic acid (UA) as the leader compound was designed to prepare a series of derivatives (three novel compounds UA-1a, UA-1b and UA-2) by modification at the C3 and C28 positions. Their chemical structures were confirmed by IR, ¹H NMR and MS. The cytotoxic activity of the derivatives was evaluated against HepG2, BGC-823 and HT-29 by the MTT assay. The novel derivative UA-1a, [3β-acetoxy-urs-12-en-28-oyl]-1-monoglyceride showed significant anti-growth ability against the assayed cancer cell lines, particularly against BGC-823, while low cytotoxicity to human normal gastric cell line GES-1. Further investigation revealed that UA-1a could induce apoptotic events of the treated BGC-823 cells, such as comet-like DNA bend, sub-G0/G1 phase accumulation and phosphatidylserine externalization. The activity of Caspase-3 was found to be up-regulated, while the expression of Bcl-2 and Survivin were down-regulated in UA-1a treated cells. UA-1a might trigger the death of BGC-823 cells by inducing apoptosis via the mitochondria pathway. UA-1a exerted stronger ability than Taxol to retard tumor growth in nude mice without leaving apparent toxicity to the hosts. The experimental data suggested that UA-1a would have a therapeutic potential in the treatment of gastric cancer.

Full text of DOI:10.1016/j.bmc.2011.05.017

Bioorganic & Medicinal Chemistry 19 (2011) 4043–4050
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Synthesis of [3b-acetoxy-urs-12-en-28-oyl]-1-monoglyceride
and investigation on its anti tumor effects against BGC-823
Kai-Kai Bai a, Fen-Ling Chen a, Zhou Yu a, Yun-Quan Zheng a, Yong-Ning Li a, Yang-Hao Guo a,b,
⇑
a College of Chemistry & Chemical Engineering, Fuzhou University, Fuzhou 350002, PR China
b Fujian Key Lab of Medical Instrument and Pharmaceutical Technology, Fuzhou 350002, PR China
a r t i c l e i n f o
a b s t r a c t
Article history:
Ursolic acid (UA) as the leader compound was designed to prepare a series of derivatives (three novel
compounds UA-1a, UA-1b and UA-2) by modification at the C3 and C28 positions. Their chemical struc-
tures were confirmed by IR, 1H NMR and MS. The cytotoxic activity of the derivatives was evaluated
against HepG2, BGC-823 and HT-29 by the MTT assay. The novel derivative UA-1a, [3b-acetoxy-urs-
12-en-28-oyl]-1-monoglyceride showed significant anti-growth ability against the assayed cancer cell
lines, particularly against BGC-823, while low cytotoxicity to human normal gastric cell line GES-1.
Further investigation revealed that UA-1a could induce apoptotic events of the treated BGC-823 cells,
such as comet-like DNA bend, sub-G0/G1 phase accumulation and phosphatidylserine externalization.
The activity of Caspase-3 was found to be up-regulated, while the expression of Bcl-2 and Survivin were
down-regulated in UA-1a treated cells. UA-1a might trigger the death of BGC-823 cells by inducing apop-
tosis via the mitochondria pathway. UA-1a exerted stronger ability than Taxol to retard tumor growth in
nude mice without leaving apparent toxicity to the hosts. The experimental data suggested that UA-1a
would have a therapeutic potential in the treatment of gastric cancer.
Received 29 March 2011
Revised 12 May 2011
Accepted 13 May 2011
Available online 23 May 2011
Keywords:
Ursolic acid
Derivative
Cytotoxicity
Apoptosis
Mitochondria pathway
Ó 2011 Elsevier Ltd. All rights reserved.
1. Introduction
In order to hunt for the potential anti-tumor drug candidates
and to study their structure-activity relationships, our group has
Current therapies for malignant epithelial tumor include che-
motherapy, surgery and radiation. Chemotherapy is widely used
for the cancer sufferers but always shows some undesirable side
effects. In order to obtain novel anticancer compounds, more and
more attention is paid to the plant kingdom. Some natural prod-
ucts have been successfully applied in clinic treatment of various
cancers. Taxol which is extracted form Taxus chinensis, has been
abroad utilized to cure malignant epithelial tumor.1
prepared series of UA derivatives. In this study a series of novel
UA derivatives were reported and their cytotoxic activities were
evaluated against human hepatoma cell HepG2, gastric cancer cell
BGC-823 and colorectal carcinoma cell HT-29. The anti-cancer
capacity of the prepared UA-1a against BGC-823 was evaluated
both in vitro and in vivo. The expression of Bcl-2 and Survivin
and the activity of Caspase-3 were determined to explore the pos-
sible anticancer mechanism.
Ursolic acid (UA, 3b-hydroxy-urs-12-en-28-oic acid), a penta-
cyclic triterpene acid existing abundantly in the plant kingdom,
has attracted the attention of the pioneers who aim to develop novel
anti tumor agents.2 Recent studies have shown that ursolic acid has
marked anti-tumor activity toward various types of cancer cell lines
in vitro and in vivo.2,3 As a prospective anticancer drug, considerable
structural modification has been performed on UA to obtain poten-
tial anticancer derivatives with enhanced pharmacokinetic/phar-
macodynamic and physical/chemical properties. Meng4 prepared
fifteen UA derivatives, some of which exerted higher anticancer
activity than UA against Hela cells. Structural modification was
carried out by Ma5 as well and some UA derivatives with enhanced
cytotoxicity upon HL-60, Bel-7402, BGC-823 and Hela cell lines
were obtained.
2. Experimental protocols
2.1. Materials
Human gastric adenocarcinoma cell line BGC-823, hepatoma cell
line HepG2, colorectal carcinoma cell line HT-29 were obtained from
Shanghai Cell Resource Center (Chinese Academy of Sciences,
Shanghai, China). Taxol and UA were purchased from Sigma
(St. Louis, MO, USA). Fetal calf serum (FCS), 3-(4,5-Dimethylthia-
zol-2-yl)-2,5-diphenyltetrazolium bromide (MTT), Penicillin/Strep-
tomycin, trypsin and RPMI-1640 medium, were provided by Gibco
(Gaithersburg, MD, USA). Caspase-3 activity detection Kit was
obtained from Beyond Time (Haimen, China). AnnexinV-FITC
apoptosis detection Kit and Cell apoptosis PI cycle detection Kit were
supplied by KeyGEN (Nanjing, China). Monoclonal anti-b-Actin,
⇑
Corresponding author. Tel./fax: +86 591 83720772.
E-mail address: yanghaoguo@yahoo.com.cn (Y.-H. Guo).
0968-0896/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2011.05.017

4044
K.-K. Bai et al. / Bioorg. Med. Chem. 19 (2011) 4043–4050
monoclonal anti-Survivin, monoclonal anti-Bcl-2 primary antibod-
ies, and horseradish peroxidase (HRP) conjugated goat anti-rabbit/
mouse IgG were from Santa Cruz Bio-technology (Inc., CA, USA).
Low melting point (LMP) agarose was a product from Invitrogen
(USA). Other reagents of an analytic grade were purchased from
commercial suppliers.
10% heat-inactivated FCS, penicillin (100 U/mL) and streptomycin
(100 g/mL) in a humidified atmosphere supplied with 5% CO2 at
37 °C. The cells in logarithmic phase were allowed for the following
l
tests.
2.4. MTT assay for cell viability/proliferation
2.2. The synthesis of UA derivatives
The cytotoxicity of the derivatives was determined by the MTT
assay.4–6 UA, Taxol and UA derivatives were dissolved in DMSO,
then diluted with culture medium. The content of DMSO in each
sample was 0.5%. Cells were seeded in 96-well plates at a density
of 5 Â 103 cells per well. After 24 h, the cells were treated with dif-
ferent concentrations of drugs for another setting period (24–72 h).
The negative control cells were treated with culture medium con-
taining 0.5% DMSO. Finally, excess MTT was added, and the culture
was incubated for another 4 h. The resultant formazan formed by
metabolically viable cells was well dissolved in 100 mL of DMSO.
The absorbance was then measured on a microplate reader appara-
tus (DNA Expert, TECAN, Switzerland) at 570 nm. Each sample had
six duplicates and the average absorbance was applied into the
following formula to calculate the growth inhibition.5 The concen-
tration of the compound, which gives the 50% growth inhibition
value, corresponds to the IC50. Each test was repeated at least three
times and the results were expressed as mean SD.
A series of ursolic acid derivatives were synthesized as shown in
Figure 1.
2.2.1. Preparation of the intermediate product (UA-3)
UA was converted to its 3-O-acetate UA-3 as previously de-
scribed.4–6 To be brief, UA (456 mg, 1 mmol) was dissolved in pyr-
idine (5 mL), mixed with acetic anhydride (0.5 mL), and the
solution was stirred at room temperature for 6 h. The reaction mix-
ture was added water (40 mL) and partitioned with dichlorometh-
ane (60 mL Â 3). The organic solution was dried over Na2SO4 and
concentrated in vacuo to give crude product. The crude product
was purified via crystallization and then recrystallization in etha-
nol to obtain UA-3.
2.2.2. General procedure for esterification at C-28 carboxylic
acid of UA and the intermediate product
UA (182 mg, 0.4 mmol) or UA-3 (200 mg, 0.4 mmol) was well
distributed (or dissolved) in 20 mL acetone, to which was added
K2CO3 (1 mmol) and KI (1 mmol). Stirred at room temperature for
6 h, the reactant was mixed with (DL)-3-Chloro-1,2-propanediol
% inhibition = [1À(OD value for treated cells/OD value for un-
treated cells)] Â 100%.
2.5. Cell cycle analysis
(100 lL, 1.2 mmol) or 2-chloroethanol (80 lL, 1.2 mmol) for an-
The cell cycle was analyzed by flow cytometry. Briefly, BGC-823
cells were treated with UA-1a (20 lM) for 24–72 h. After the incu-
other 24 h. The mixture was concentrated in vacuo, mixed with dis-
tilled water, neutralized with HCl, filtered to get crude solid. The
crude was purified by chromatography on a silica gel column eluted
with petroleumether/ethylacetate/triethylamine (V/V/V = 33:66:1)
to get purified compounds UA-1a, UA-1b and UA-2.
bation, approximately 106 treated cells were harvested, and the
cell cycle phase distribution was assessed by propidium iodide
(PI) staining, according to the manufacturer’s instruction (Cell
apoptosis PI cycle detection Kit) in a flow cytometer (Beckman
Coulter, EPICS XL, USA). The percentage of cells in G0/G1, S,
G2/M and sub-G0/G1 phase were analyzed by means of CellQuest
software program.7 Similar process was performed after the incu-
2.2.3. Determination of the physical/chemical properties of UA
derivatives
The melting points were determined on an electrically heated
X-4 digital visual melting point apparatus and are uncorrected. IR
spectra were recorded on Thermo Nicolet Nexus 470 spectrometer.
1H NMR spectra were recorded on a BRUKER AVANCE 600 MHz
spectrometer at room temperature, and chemical shifts were mea-
sured in ppm downfield from TMS as internal standard. Mass spec-
tra were recorded on Agilent 1100 LC/MSD Trap XCT equipment, in
positive Electron Spray Ionization (ESI) mode and were reported as
m/z. UV–vis absorbance spectra were recorded on a Shimadzu UV-
1700 Pharma spectrometer.
bation with different concentration of UA-1a (10, 20, 40 lM) for
48 h to determine the effects of the dosage on the DNA content.
All experiments were repeated at least three times.
2.6. Annexin-V/PI dual-staining assay
Tumor cells were treated with 0, 10, 20, 40 lM UA-1a for 48 h,
washed and resuspended in PBS buffer. Apoptotic cells were
identified by double staining with recombinant fluorescein isothio-
cyanate (FITC)-conjugated Annexin-V and PI8, by using the Annex-
inV-FITC apoptosis detection Kit following the manufacturer’s
instructions. Flow cytometric analysis was performed immediately
after staining. Data acquisition and analysis were performed by
using CellQuest software.7
2.3. Cell line and culture
The human cell lines including BGC-823, HepG2, HT-29 and
GES-1 were cultured in RPMI-1640 medium supplemented with
R
R
H
H
COOH
H
COOH
a
b
H
b
O
or c
O
CH3COO
HO
CH3COO
HO
H
H
H
H
UA-1a , R=OCH2CH(OH)CH2OH
UA-1b , R=OCH2CH2OH
UA
UA-2 , R=OCH2CH(OH)CH2OH
UA-3
Figure 1. Route for synthesis of ursolic acid derivatives. Reagents and conditions: (a) (CH3CO)2O, pyridine, at room temperature; (b) (DL)-3-chloro-1,2-propanediol, K2CO3,
acetone, reflux; (c) 2-chloroethanol, K2CO3, acetone, reflux.

K.-K. Bai et al. / Bioorg. Med. Chem. 19 (2011) 4043–4050
4045
2.7. Comet assay
experiments). Results with a P value less than 0.05 were considered
to be statistically significant.
BGC-823 cells were incubated for 48 h at 37 °C in the presence
of UA-1a (10, 20, 40 l
M). After the treatment, about 105 treated
3. Results
cells were harvested to perform the alkaline comet assay as de-
scribed by Thierry Godard.9 The comet bends were recorded by
GelPro Documentation System (Biosens 810, Shanghai, China).
3.1. Properties of UA derivatives
UA-3, UA-1a, UA-1b and UA-2 with high purity were prepared
as described in Section 2.2. Their physical/chemical properties
were as followed:
2.8. Determination of Total Caspase-3 enzyme activity
BGC-823 cells were treated with UA-1a (0, 10, 20 lM) for 48 h.
After the treatments cells were detached with trypsin/EDTA and
washed twice with cold PBS buffer. Then, about 106 cells were har-
vested and the total Caspase-3 activities were determined by using
the Caspase-3 activity detection Kit following the manufacturer’s
instructions. Briefly, the collected cells were lysed and the lysis
was incubated with sufficient Ac-DEVD-pNA (acetyl-Asp-Glu-Val-
Asp p-nitroanilide, the corresponding fluorogenic substrate) for
1 h. Cleavage of substrate was monitored by DNA expert (TECAN,-
CA) at 405 nm. Additionally, Bradford Protein Assay10,11 was used
to quantify the content of total protein. Results were expressed
as the change in fluorescence units (per 1 g of protein) relative to
control.
3.1.1. 3b-Acetoxy-urs-12-en-28-oic acid (UA-3)
While crystalline powder; yield 70.5%; mp: 284–287 °C (lit.14,15
285 °C, 287–289 °C); IR (KBr): 3366, 2972, 2928, 1736, 1697, 1455,
1370, 1246 cmÀ1; ESI-MS: m/z 499.6 [M+H]+. UV–vis (methanol),
kmax = 210 nm; 1H NMR (600 MHz, CDCl3): 5.32 (t, 1H, J = 3.6 Hz,
H-12), 4.42 (t-like, 1H, H-3), 2.22 (d, 1H, J = 11.2 Hz, H-18), 2.04
(s, 3H, CH3CO), 0.92 (d, 3H, J = 6.4 Hz, CH3), 0.85 (d, 3H, J = 6.6 Hz,
CH3), 1.09, 1.00, 0.91, 0.78, 0.74 (s, 15H, 5Â CH3).
3.1.2. [3b-Acetoxy-urs-12-en-28-oyl]-1-monoglyceride (UA-1a)
While crystalline powder; yield 85.8%; mp: 127–129 °C; IR
(KBr): 3442, 2970, 2928, 2874, 1735, 1245 cmÀ1; ESI-MS: m/z
595.5 [M+Na]+. UV–vis (Methanol), kmax = 209 nm; 1H NMR
(600 MHz, CDCl3): 5.25 (t, 1H, J = 3.6 Hz, H-12), 4.51–4.48 (m, 1H,
H-3), 4.12–4.06 (m, 1H, COOCH2a), 4.20–4.17 (dd, 0.5H,
J = 11.5,4.5 Hz, COOCH2b), 4.04–4.01 (dd, 0.5H, J = 11.5,6.0 Hz,
COOCH2b), 3.95–3.91 (m, 0.5H, CH(OH)CH2OH), 3.91–3.87 (m,
0.5H, CH(OH)CH2OH), 3.70–3.67 (m, 1H, HOCH2a), 3.60–3.57 (m,
1H, HOCH2b), 2.22 (d, 1H, J = 11.2 Hz, H-18), 2.04 (s, 3H, CH3CO),
0.95 (d, 3H, J = 6.2 Hz, CH3), 0.87 (d, 3H, J = 6.6 Hz, CH3), 1.08,
0.94, 0.86, 0.85, 0.75 (s, 15H, 5Â CH3).
2.9. Western blot analysis
After the incubation with UA-1a (10, 20, 40 lM) for 48 h, the
BGC-823 cells were detached with trypsin/EDTA. The determina-
tion of Bcl-2 and Survivin proteins was performed as described.12,13
To be brief, more than 106 cells were lysed and the total protein
was harvested and quantified via Bradford Protein Assay.10,11 For
each sample, 150–200
SDS–polyacrylamide gel, electrophoresed and transferred to a
PVDF membrane (0.45 m). The primary antibodies were used at
lg of protein was loaded onto 12.5%
l
3.1.3. [3b-Hydroxy-urs-12-en-28-oyl]-1-monoglyceride (UA-2)
While crystalline powder; yield 86.0%; mp: 100–102 °C; IR
a concentration of 1:1000 for Survivin, 1:200 for Bcl-2 and
1:1000 for b-actin. The membranes were incubated for 1 h with
HRP-labeled secondary antibodies, then developed by an ECL
system following the manufacturer’s instructions (Beyotime, Nan-
jing, China). Quantification of band intensity was performed by
means of Gel-Pro Analyzer 3.1 software.12
(KBr): 3351, 2948, 2925, 2869, 1788, 1718, 1456, 1184,1044 cm
À1
;
UV–vis (methanol), kmax = 207 nm; ESI-MS: m/z 553.5
[M+Na]+; 1H NMR (600 MHz, CDCl3): 5.26 (t, 1H, J = 3.6 Hz, H-12),
4.12–4.06 (m, 1H, COOCH2a), 4.20–4.17 (dd, 0.5H, J = 11.5,4.5 Hz,
COOCH2b), 4.04–4.01 (dd, 0.5H, J = 11.5,6.0 Hz, COOCH2b),
3.94–3.91 (m, 0.5H, CH(OH)CH2OH), 3.90–3.87 (m, 0.5H, CH(OH)-
CH2OH), 3.70–3.67 (m, 1H, HOCH2a), 3.60–3.57 (m, 1H, HOCH2b),
3.22–3.20 (dd, 1H, J = 11.3,4.5 Hz, H-3), 2.22 (d, 1H, J = 11.2 Hz,
H-18), 0.95 (d, 3H, J = 6.2 Hz, CH3), 0.85 (d, 3H, J = 6.4 Hz, CH3),
1.09, 0.99, 0.91, 0.78, 0.75 (s, 15H, 5Â CH3). .
2.10. Anti-tumor effect in vivo
The in vivo therapeutic effect of UA-1a was determined as de-
scribed by Zheng.8 Male nude mice (5 weeks old, 18–22 g body
weight) were subcutaneously inoculated with 1 Â 107 BGC-823
cells. After tumor volume reached about 200 mm3, the tumor-bear-
ing nude mice were randomized into four experimental groups,
each with eight mice. The control mice were given with 0.2 mL
0.9% saline each. Two treatment groups were intravenously injected
with 0.2 mL UA-1a solution at the doses of 6 mg/Kg and 30 mg/Kg,
respectively. In addition, the positive control mice were given with
0.2 mL Taxol solution at the dose of 8 mg/Kg. The tumor volume
were calculated once a day using the following formula: tumor vol-
ume = 0.5  (length  width  width). After tail vein injections for
five times (day 1st, 4th, 7th, 10th, and 13th), all the animals were
sacrificed at day 16th and the tumors were dissected and weighted.
3.1.4. [3b-Acetoxy-urs-12-en-28-oyl]-1-ethylene gylcol
monoester (UA-1b)
While crystalline powder; yield 90.1%; mp: 159–161 °C; IR
(KBr): 3511, 2970, 2948, 2930, 2878, 1731, 1694, 1455, 1371,
1243 cmÀ1; UV–vis (methanol), kmax = 212 nm; ESI-MS: m/z 565.6
[M+Na]+; 1H NMR(600 MHz, CDCl3): 5.26 (t, 1H, J = 3.6 Hz, H-12),
4.51–4.48 (m, 1H, H-3), 4.22–4.18 (t, 1H, COOCH2a), 4.10–4.06
(m, 1H, COOCH2b), 3.82–3.75 (m, 2H, HOCH2), 2.25 (d, 1H,
J = 11.2 Hz, H-18), 2.04 (s, 3H, CH3CO), 0.95 (d, 3H, J = 6.2 Hz,
CH3), 0.87 (d, 3H, J = 6.6 Hz, CH3), 1.09, 0.94, 0.86, 0.85, 0.77 (s,
15H, 5Â CH3);
2.11. Statistic analysis
3.2. Cytotoxicity of the UA derivatives against different cell lines
Results were expressed as the mean standard deviation (SD) of
data obtained from at least triplicate experiments. A statistical anal-
ysis was performed with one-way analysis of variance (ANOVA)
using SPSS 13.0 analytical software. Data were expressed as
mean SD (n P 3, where n represents the number of independent
The in vitro cytotoxic activity was evaluated for UA and its deriv-
atives against three tumor cell lines (HepG2, HT-29, BGC-823) and a
human normal gastric cell line (GES-1). Taxol was used as a positive
control. The results were shown in Table 1. It can be seen from
Table 1 that UA suppressed proliferation of the three assayed cancer

4046
K.-K. Bai et al. / Bioorg. Med. Chem. 19 (2011) 4043–4050
cells in different extents (IC50 values of 23.0–53.4
lM) and exhibited
some toxic effect on GES-1 (IC50 value of 93.8
lM). The prepared UA
derivatives exhibited more effective anti cancer activity. Introduc-
tion of a monoglyceride at the C-28 position, giving birth to UA-2,
might obviously reduce the toxicity for GES-1 cells. UA-3, the 3-O-
acetate of UA, demonstrated higher cytotoxicity than UA against
the three cancer cell lines. UA-1a and UA-1b were two compounds
resulting from the esterification of the carboxyl group at the C-28
position of UA-3 with glycerol and ethylene glycol, respectively.
Compared with UA, both UA-1a and UA-1b showed the higher anti
cancer activity and the lower cytotoxicity for the normal GES-1cells.
UA-1a showed the most significant antiproliferative ability
among the synthesized UA derivatives. It possessed stronger
anti-tumor activity against BGC-823 cells than Taxol. The IC50
Figure 2. Cytotoxicity of UA-1a against human gastric carcinoma cell line BGC-823.
Growth inhibition was determined by MTT assay. Each value represented the
mean SD of three independent experiments, each of which had six replicates.
value of UA-1a was 15.66 lM after treating BGC-823 cells for
48 h, which was 50% lower than that of Taxol (Table 1). This UA
derivative inhibited the growth of BGC-823 cells in a dose- and
time-dependent manner (Fig. 2). The suppression of cells growth
was rapidly heightened when treated with UA-1a at the concentra-
more defined heads were scored as normal cells with minimal
DNA diffusion, whereas cells with diffuse fan-like tails and small
heads were considered as apoptotic cells. It was clearly demon-
strated that the capacity of UA-1a to trigger apoptosis of
BGC-823 cells was dependent on the dose.
tions ranged from 10
obvious morphological changes of BGC-823 cells were observed
after the treatment with UA-1a for 48 h at 30 M (data not shown).
lM to 30 lM. Cell shrinkages and other
l
3.5. Annexin-V/PI dual-staining assay of BGC-823 cells treated
by UA-1a
In addition, UA-1a had lower toxicity than Taxol against GES-1
cells (Table 1).
The novel UA derivative UA-1a might have a therapeutic poten-
tial in the treatment of gastric tumor since it showed more signifi-
cant anti-proliferation activity against BGC-823 cells and greater
safety for the normal cells than UA and Taxol. The anti-cancer mech-
anism and the in vivo anti-tumor effects needed further research.
The apoptotic effect of UA-1a was further evaluated by Annex-
inV-FITC/PI (AV/PI) dual staining experiment to examine the occur-
rence of phosphatidylserine externalization onto the cell surface.
This experiment allowed the differentiation among normal alive
cells (AVÀ/PIÀ), early-phase apoptotic cells (AV+/PIÀ), and late-
phase apoptotic or necrotic cells (AV+/PI+). As shown in Figure 5,
when the cells were not treated with UA-1a, as a control, 95.5% of
cells were in the normal condition. When the dose of UA-1a in-
3.3. Cell cycle distribution of BGC-823 cells treated by UA-1a
To characterize the cell death triggered by UA-1a, cell cycle anal-
ysis was performed by staining the cells with PI. The results were
shown in Figure 3. Compared with the negative control, a dose-
dependent increasing in the percentage of cells in sub-G0/G1 phase
was observed (from 0.66% in control cells to 96.13% when treated
by 40 lM UA-1a) (Fig. 3A). Meanwhile, the percentage of cells in
sub-G0/G1 phase was increased in a time-dependent manner, with
3.71%, 7.91% and 13.63% of sub-G0/G1 cells for 24 h, 48 h, and 72 h,
respectively when treated by 20 lM UA-1a (Fig. 3B). Treatment of
BGC-823 cells with UA-1a distinctly resulted in a dose- and time-
dependent apoptosis. The cell cycle distribution (such as G0/G1, S,
G2/M) was slightly modified under different treating conditions
and no marked arrest at any phase of the cell cycle was found.
creased up to 20 lM, the population of normal cells was decreased
to 72.2%, however 19.4% of treated cells entered into the early apop-
totic stage. A substantial amount of apoptotic cells, 78.8% both in
the early apoptotic stage and in the late apoptotic stage, were de-
tected after being treated by UA-1a at 30 lM. Treating with higher
concentration of UA-1a induced a shift of the cell population to the
late apoptotic stage. 89.5% of cells were observed in the late apop-
totic/necrotic phase when treated with 40 lM of UA-1a for 48 h.
The data indicated that the population of apoptotic cells increased
in a dose-dependent manner and the critical concentration for the
induction of apoptosis of BGC-823 by UA-1a was about 30 lM.
3.6. Determination of Caspase-3 activity in BGC-823 cells treated
by UA-1a
3.4. Comet assay of BGC-823 cells treated by UA-1a
The proteolytic activity of the executioner Caspase-3 was deter-
mined to explore the role of caspases in the apoptosis induced by
UA-1a. As shown in Figure 6, a significant increase of Caspase-3
activity in the BGC-823 cells was observed after 48 h exposure to
UA-1a. Compared with the control, the total Caspase-3 activity
was increased up to 2.54-fold and 3.65-fold, respectively after
The comet assay is a sensitive method for detection of DNA
fragmentation occurring during apoptosis.9 The comet effect of
UA-1a on BGC-823 cells was shown in Figure 4. Cells with larger,
Table 1
The in vitro cytotoxicity of ursolic acid (UA) derivatives [expressed as IC50 (lM)]
against HT-29, HepG2 and BGC-823 human cancer cell lines, as well as the normal
human gastric epithelial cell line GES-1 (48 h)
BGC-823 cells were separately incubated with 10 lM and 20 lM
of UA-1a. It was probable that UA-1a induced the apoptosis of
BGC-823 cells via the caspase-dependent pathways mediated by
the activation of Caspase-3.
Compound HT-29
HepG2
BGC-823
GES-1
Taxol
UA
UA-2a
UA-3
UA-1a
UA-1b
<10
30.74 4.59
53.42 4.07
33.09 5.91
22.96 0.99
106.2 13.1
93.83 8.51
154.3 19.2**
92.94 6.34
26.31 2.74
ntb
44.23 2.81** 20.51 4.26
37.67 2.92** 19.36 1.70*
3.7. Effect of UA-1a on the expression of Survivin and Bcl-2
23.89 1.52*
18.43 1.83** 27.46 1.78** 15.66 1.38**,a 155.4 11.4**,a
23.56 1.69*
29.83 3.34** 17.22 2.50**
137.2 10.9**
Some special proteins have been reported to be associated with
the apoptosis pathway. In this work the effect of UA-1a on the
expression of Survivin and Bcl-2 proteins was examined (Fig. 7A).
The Survivin expression in cells treated with UA-1a (10, 20,
*
P < 0.05 versus UA.
P < 0.01 versus UA.
**
a
P < 0.01 versus Taxol.
b
40 lM) were decreased by 37.6%, 43.1% and 53.1%, respectively
Not tested.

K.-K. Bai et al. / Bioorg. Med. Chem. 19 (2011) 4043–4050
4047
Figure 3. The cell cycle distribution of BGC-823 cells treated with (A) 0, 10, 20, 40
lM of UA-1a for 48 h and with (B) 20
lM of UA-1a for 0, 24, 48, 72 h was shown. The
percentage of cells residing in sub-G0/G1phase was counted. Simultaneously, the cell population of G0/G1, S and G2/M phases that constituted the DNA-integrity cells (taken
⁄
as 100%) were analysed. Results were expressed as mean SD. P < 0.05; ⁄⁄P < 0.01, versus control (0
lM)
Control
UA-1a (10µM, 48 h)
UA-1a (20µM, 48 h)
UA-1a (40µM, 48 h)
Figure 4. Comet-like DNA electrophoresis bends in BGC-823 cells treated with UA-1a at different concentrations (0, 10, 20 and 40 lM) for 48 h.
when compared to that of the control (Fig. 7B). In addition, UA-1a
significantly down-regulated the expression of Bcl-2 (Fig. 7B). It
was evident that UA-1a suppressed the expression of Survivin
and Bcl-2 in a dose-dependent manner. The results indicated that
the inhibition of expression of Survivin and bcl-2 genes was in-
volved in the anti-tumor activity of UA-1a.
tural modification has been performed at the C-3 or C-28 position of
UA to enhance its pharmacokinetic /pharmacodynamic properties.
The results of Ma5 revealed that the triterpenes possessing two
hydrogen-bond forming groups (an H-donor and a carbonyl group)
at positions C-3 and C-28 exhibit cytotoxic activity. Introduction of
an amino group at these two positions might increase the cytotoxic-
ity. Meng4 concluded that both the 3-O-acetyl group and a 28-amido
group might be important for improving the anti tumor activity,
however a hydroxyl group at the C-28 amide side chain could not
influence the antiproliferative effect.
In this study, some UA derivatives were prepared by esterifica-
tion at C-3 or/and C-28 positions of the parent ursolic acid. The
structure-activity relationships (SAR) of these UA derivatives sug-
gested the 3-O-acetyl group might be important for improving
the anti-tumor activity. The introduction of a monoglyceride or a
ethylene gylcol monoester at the C-28 position could reduce the
toxicity of ursolic acid or its 3-O-acetate toward the normal gastric
cells.
UA-1a was synthesized via condensation of acetic acid with the
hydroxyl group at the C-3 position of ursolic acid and then by
esterification of the carboxyl group at C-28 position with glycerol.
This novel compound showed strengthened anti-cancer effect. The
reason was probably contributed to the obtained better lipophilic-
ity, which often resulted in better membrane permeability, fol-
lowed with better pharmaceutical effects.16 The experimental
data also revealed the esterification of the carbonyl group of the
mother nucleus might be important for the cytotoxicity. The intro-
duction of polylol to the C-28 position could benefit both of the
anti-tumor activity and the safety toward normal cells.
3.8. In vivo anti-tumor effect of UA-1a
The above experiments indicated UA-1a exhibited an anti-pro-
liferation effect on the BGC-823 cells in vitro. Subsequently, a kind
of animal model, nude mice subcutaneously injected with BGC-823
cells was employed to explore if UA-1a was able to suppress the
tumor growth in vivo. It was found that the tumor sizes in the mice
treated by UA-1a were significantly smaller than that of the mice
in the control group. After five times of tail vein injection, the tu-
mor volume were diminished by 29.3% (at low dose, 6 mg/Kg)
and 41.9% (at high dose, 30 mg/Kg), when compared with the con-
trol (Fig. 8A). The solid tumors of each group were weighted and
the data showed that the tumors treated with UA-1a were mark-
edly lighter than that without any treatment (Fig. 8B).
Taxol was used as a positive control. The determined tumor vol-
ume was decreased by 15.2% when treated with Taxol at the dose
of 8 mg/Kg (Fig. 8 A&B). We concluded UA-1a exerted stronger
ability to inhibit tumor growth than Taxol. Moreover, no significant
difference was observed on the body weight of nude mice treated
with UA-1a when compared with the control group (Table 2),
which demonstrated that UA-1a retarded tumor growth in nude
mice without leaving apparent toxicity to the hosts.
Apoptosis, or programmed cell death, plays an essential role in
developing and maintaining tissue homeostasis through eliminating
genetically damaged cells and excess cells.12 It is a fundamental pro-
cess involving the degradation of DNA in a scheduled manner.17 After
UA-1a treatment, the apoptosis events of the treated BGC-823 cells,
4. Discussion
Ursolic acid (UA) has been shown to possess anti-proliferative
ability against various types of cancer cells.2–6 More and more struc-

4048
K.-K. Bai et al. / Bioorg. Med. Chem. 19 (2011) 4043–4050
Figure 5. Annexin V/PI dual staining of BGC-823 cells treated with UA-1a (0, 10, 20, 30 and 40 lM) for 48 h was carried out. The cells were harvested, stained and then
analyzed by Flow cytometry. In all panels, cells in the lower left quadrant (M3: AVÀ/PIÀ) were alive, cells in the lower right quadrant (M4: AV+/PIÀ) were in early apoptosis,
cells in the upper right quadrant (M2: AV+/PI+) were in late apoptosis/necrosis, and cells in the upper left quadrant (M1: AVÀ/PI+) were damaged appearing in the process of
cell collection. Percentage of total signal within the quadrant was indicated.
endonulease responsible for DNA degradation as well as forming
of comet-like DNA bend.18,19 It was found in our experiments that
the activity of caspase-3 could be enhanced more than 2 times by
UA-1a. These results suggested that UA-1a trigger the cancer cell
death by caspases-dependent apoptotic mechanism involving the
activation of caspase-3.
Gastric carcinoma exhibits high level of Bcl-2 protein expression.
Bcl-2 overexpression protects cancer cells from apoptosis induced
by different stimuli.20 In this work, down-regulation of Bcl-2
expression by UA-1a initiated apoptosis of BGC-823 cells in a casp-
ases-dependent apoptotic pathway mediated by Bcl-2 protein.
Moreover, Bcl-2 is a member of the Bcl-2 family proteins which
provided the link between the caspase cascade and the mitochon-
Figure 6. The relative Caspase-3 activity in BGC-823 cells treated by UA-1a (0, 10,
20 lM). Enzymatic activities of Caspase-3 were defined as the fluorescence per one
dria.12,21 It was probable that the programmed cell death of BGC-
gram of protein in the treated sample. Relative Caspase-3 activity was shown as the
ratio of enzymatic activities of UA-1a treated cells in respect to that of the control
823 cells treated by UA-1a proceeded in the mitochondria pathway.
Survivin is a member of the inhibitors of apoptosis protein fam-
ily (IAP).13 A sustained overexpression of Survivin is a characteris-
tic feature of gastric cancer, and plays an important role in tumor
progression by inhibiting apoptosis and facilitating mitosis, giving
cancer cells a survival and growth advantage.22 In this work, the
expression of Survivin was greatly inhibited by UA-1a in a dose-
dependent manner. The down-regulation of Survivin expression
resulted in the increase in the population of apoptotic cells.
In conclusion, our experimental data exhibited a satisfactory
anti-cancer activity of UA-1a against BGC-823 cells and suggested
a clear anti-cancer mechanism. It was notable that UA-1a exerted
stronger ability than Taxol to retard tumor growth in nude mice
(0 lM). Results were express as mean SD of three independent experiments.
⁄P<0.05;⁄⁄P<0.01 versus control.
including comet-like DNA bend and sub-G0/G1 phase were observed.
Annexin V/PI dual staining assay further provided an explicit evi-
dence for the induction-effect of UA-1a on the apoptosis of BGC-
823 cells. The experimental data suggested that apoptosis play an
important role in the proliferation inhibition of BGC-823 cells by
UA-1a.
Caspase-3 is a key apoptotic executive caspase, known to
mainly mediate the caspase activated DNase (CAD) which is an

K.-K. Bai et al. / Bioorg. Med. Chem. 19 (2011) 4043–4050
4049
Figure 7. Effects of various concentrations of UA-1a on the expression of Survivin and Bcl-2 proteins in BGC-823 cells. (A) Western blot analysis of Survivin and Bcl-2;
(B) Quantification of relative expression of Survivin and Bcl-2 when comparing with the control.
Figure 8. Anti-tumor effects of UA-1a on nude mice bearing BGC-823 cells. (A) The relative tumor volume and (B) the tumor weight were shown as mean SD of eight
replicates. ⁄P<0.05; ⁄⁄P<0.01 versus control.
Table 2
Effects of UA-1a on the body weight (gramme) of BGC-823 cells xenograft nude mice (n = 8)
Days of treatment
Control
Taxol (8 mg/Kg)
UA-1a (6 mg/Kg)
UA-1a (30 mg/Kg)
1
4
7
10
13
16
19.69 1.21
20.58 1.35
20.79 1.29
20.54 1.45
20.21 1.83
20.12 1.52
18.76 1.49
19.72 0.92
19.70 1.10
20.13 1.37
20.17 1.61
20.25 1.43
19.21 0.82
20.66 1.16
20.13 1.20
20.03 1.18
20.07 1.14
20.09 1.25
19.26 1.16
19.61 1.04
19.29 0.92
19.16 2.23
19.69 2.02
19.57 1.78
3. Yang, L.; Liu, X.-Z.; Lu, Z.-B.; Chan, J. Y.-W.; Zhou, L.-L.; Fung, K. P.; Wu, P.; Wu,
S.-H. Cancer Lett. 2010, 298, 128.
4. Meng, Y.-Q.; Liu, D.; Cai, L.-L.; Chen, H.; Cao, H.; Wang, Y.-Z. Bioorg. Med. Chem.
2009, 17, 848.
without leaving apparent toxicity to the hosts. UA-1a might have a
therapeutic potential in the treatment of gastric cancer.
5. Ma, C.-M.; Cai, S.-Q.; Cui, J.-R.; Wang, R.-Q.; Tu, P.-F.; Masao, H.; Mohsen, D. Eur.
J. Med. Chem. 2005, 40, 582.
Acknowledgments
6. Tu, H.-Y.; Huang, A.-M.; Wei, B.-L.; Gan, K.-H.; Hour, T. C.; Yang, S.-C.; Pu, Y.-S.;
Lin, C.-N. Bioorg. Med. Chem. 2009, 17, 7265.
7. Li, Y.-C.; Qu, X.-J.; Qu, J.-L.; Zhang, Y.; Liu, J.; Liu, Y.-P. Cancer Lett. 2009, 284,
208.
8. Zheng, Y. Q.; Xin, Y. W.; Shi, X. A.; Guo, Y. H. Appl. Microbiol. Biotechnol. 2010, 88,
1169.
The authors are grateful for funds provided by Natural Science
Foundation of Fujian Province of China (No.2008J1005) and
Open-foundation of the Fujian Key Lab of Medical Instrument
and Pharmaceutical Technology, China (No.09003).
9. Godard, T.; Deslandes, E.; Sichel, F.; Poul, J. M.; Gauduchon, P. Mutat. Res. 2002,
520, 47.
10. Noble, J. E.; Bailey, Marc. Method Enzymol. 2009, 463, 73.
11. Barbosa, H.; Slater, N. K. H.; Marcos, J. C. Anal. biochem. 2009, 395, 108.
12. Sun, B.; Geng, S.; Huang, X.-J.; Zhu, J.; Liu, S.; Zhang, Y.-J.; Yue, J.; Li, Y.-J.; Wang,
J.-Z. Cancer Lett. 2010, 301, 95.
13. Chen, T.; Deng, C. _S. Int. Immunopharmacol. 2008, 8, 1006.
14. Lassak, E. V.; Pinhey, J. T. Aust. J. Chem. 1968, 21, 1927.
15. Gopinath, K. W.; Mohamed, P. A.; Kidwai, A. R. Indian J. Chem. 1963, 1, 98.
16. Stella, V. J.; Nti-Addae, K. W. Adv. Drug Del. Rev. 2007, 59, 677.
17. Qiao, L.; Hanif, R.; Sphicas, E.; Shifff, S. J.; Riga, B. Biochem. Pharmacol. 1998, 55,
53.
A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmc.2011.05.017.
References and notes
1. Malik, S.; Cusidó, R. M.; Mirjalili, M. H.; Moyano, E.; Palazón, J.; Bonfill, M.
Process Biochem. 2011, 46(1), 23.
2. Liu, J. J. Ethnopharmacol. 2005, 100, 92.
18. Krantic, S.; Mechawar, N.; Reix, S.; Quirion, R. Prog. Neurobiol. 2007, 81, 179.

4050
K.-K. Bai et al. / Bioorg. Med. Chem. 19 (2011) 4043–4050
19. Thippeswamy, G.; Salimath, B. P. Environ. Toxicol. Phar. 2007, 23, 212.
20. Du, Y.-L.; Yin, F.; Liu, C.-J.; Hu, S.-J.; Wang, J.; Xie, H.-H.; Hong, L.; Fan, D.-M.
Biochem. Biophys. Res. Commun. 2006, 345, 1092.
21. Dia, V. P.; de Mejia, E. G. Cancer Lett. 2010, 295, 44.
22. Tarnawski, A.; Pai, R.; Chiou, S.-K.; Chai, J.; Chu, E. C. Biochem. Biophys. Res.
Commun. 2005, 334, 207.