Structural analysis of isosteviol and related compounds as DNA polymerase and DNA topoisomerase inhibitors

Article abstract of DOI:10.1016/j.lfs.2005.03.022

Isosteviol (ent-16-ketobeyeran-19-oic acid) is a hydrolysis product of stevioside, which is a natural sweetener produced in the leaves of Stevia rebaudiana (Bertoni) Bertoni. In this report, we prepared isosteviol and related compounds from stevioside by microbial transformation and chemical conversion and assayed the inhibitory activities toward DNA metabolic enzymes and human cancer cell growth. Among twelve compounds obtained, only isosteviol (compound 3) potently inhibited both mammalian DNA polymerases (pols) and human DNA topoisomerase II (topo II), and IC₅₀ value for pol α was 64.0 μM. This compound had no inhibitory effect on higher plant (cauliflower) pols, prokaryotic pols, human topo I, and DNA metabolic enzymes such as human telomerase, T7 RNA polymerase, and bovine deoxyribonuclease I. With pol α, isosteviol acted non-competitively with the DNA template-primer and nucleotide substrate. Isosteviol prevented the growth of human cancer cells, with LD ₅₀ values of 84-167 μM, and 500 μg of the compound caused a marked reduction in TPA (12-O-tetradecanoylphorbol-13-acetate)-induced inflammation (inhibitory effect, 53.0%). The relationship between the structure of stevioside-based compounds and these activities were discussed.

Full text of DOI:10.1016/j.lfs.2005.03.022

Life Sciences 77 (2005) 2127–2140
www.elsevier.com/locate/lifescie
Structural analysis of isosteviol and related compounds as
DNA polymerase and DNA topoisomerase inhibitors
a,b,
c
c
c
, Toshihiro Akihisa , Motohiko Ukiya , Yusuke Hamasaki ,
T
Yoshiyuki Mizushina
Chikako Murakami-Nakai , Isoko Kuriyama , Toshifumi Takeuchi ,
a,1
a
d
d
Fumio Sugawara , Hiromi Yoshida
a,b
a
Laboratory of Food and Nutritional Sciences, Department of Nutritional Science, Kobe-Gakuin University, Nishi-ku,
Kobe, Hyogo 651-2180, Japan
b
High Technology Research Center, Kobe-Gakuin University, Nishi-ku, Kobe, Hyogo 651-2180, Japan
c
College of Science and Technology, Nihon University, 1-8 Kanda Surugadai, Chiyoda-ku, Tokyo 101-8308, Japan
d
Department of Applied Biological Science, Tokyo University of Science, Noda, Chiba 278-8510, Japan
Received 17 January 2005; accepted 3 March 2005
Abstract
Isosteviol (ent-16-ketobeyeran-19-oic acid) is a hydrolysis product of stevioside, which is a natural
sweetener produced in the leaves of Stevia rebaudiana (Bertoni) Bertoni. In this report, we prepared isosteviol
and related compounds from stevioside by microbial transformation and chemical conversion and assayed the
inhibitory activities toward DNA metabolic enzymes and human cancer cell growth. Among twelve
compounds obtained, only isosteviol (compound 3) potently inhibited both mammalian DNA polymerases
(pols) and human DNA topoisomerase II (topo II), and IC50 value for pol a was 64.0 AM. This compound had
no inhibitory effect on higher plant (cauliflower) pols, prokaryotic pols, human topo I, and DNA metabolic
Abbreviations: topo, DNA topoisomerase; pol, DNA polymerase; dTTP, 2V-deoxythymidine 5V-triphosphate; dNTP,
deoxyribonucleoside triphosphates; DMSO, dimethylsulfoxide; dsDNA, double stranded DNA; EtBr, ethidium bromide; MTT,
3
-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide; TPA, 12-O-tetradecanoylphorbol-13-acetate.
T Corresponding author. Laboratory of Food and Nutritional Sciences, Department of Nutritional Science, Kobe-Gakuin
University, Nishi-ku, Kobe, Hyogo 651-2180, Japan. Tel.: +81 78 974 1551x3232; fax: +81 78 974 5689.
E-mail address: mizushin@nutr.kobegakuin.ac.jp (Y. Mizushina).
Present address: Department of Intractable Diseases, International Medical Center of Japan, Shinjuku-ku, Tokyo 162-
1
8
655, Japan.
0
024-3205/$ - see front matter D 2005 Elsevier Inc. All rights reserved.
doi:10.1016/j.lfs.2005.03.022

2
128
Y. Mizushina et al. / Life Sciences 77 (2005) 2127–2140
enzymes such as human telomerase, T7 RNA polymerase, and bovine deoxyribonuclease I. With pol a,
isosteviol acted non-competitively with the DNA template-primer and nucleotide substrate. Isosteviol prevented
the growth of human cancer cells, with LD50 values of 84–167 AM, and 500 Ag of the compound caused a
marked reduction in TPA (12-O-tetradecanoylphorbol-13-acetate)-induced inflammation (inhibitory effect,
53.0%). The relationship between the structure of stevioside-based compounds and these activities were
discussed.
D 2005 Elsevier Inc. All rights reserved.
Keywords: Isosteviol; DNA polymerase; DNA topoisomerase; Enzyme inhibitor; Cytotoxicity; Anti-inflammation
Introduction
Stevioside (compound 1) is a glycoside formed by three glucose molecules and steviol
compound 2), a diterpenic carboxylic alcohol (Fig. 1). Stevioside is a high-intensity sweetener that
(
tastes about 300 times sweeter than sucrose (0.4% solution). In many countries, it is used as a low-
calorie sweetener in a wide range of food products and beverages. Stevioside has been used for
several years as a sweetener in South America, Asia, Japan, China, and different countries of the
European Union, while in the United States, it has been used as a dietary supplement since 1995.
Stevia rebaudiana (Bertoni) Bertoni products are approved for sweetening purposes in Brazil,
Korea, and Japan (Kinghorn, 2002). The advantages of stevioside as a dietary supplement for
human subjects are manifold: it is stable, it is non-calorific, it maintains dental health by reducing
the intake of sugar, and it has possibilities for use by diabetic and phenylketonuria patients and
obese persons.
We have reported that novel anti-TPA-induced inflammatory compounds, a novel terpeno benzoic
acid (i.e., myrsinoic acid A (MAA)), novel triterpenoids (i.e., tormentic acid (TA) and euscaphic acid
(EA)) and some phenolic compounds such as curcumin were inhibitors of mammalian DNA
polymerase a, h and E (pol a, h and E) (Mizushina et al., 2000, 2003; Murakami et al., 2002).
Although tumor promoters are compounds that promote tumor formation (Hecker, 1978), these
compounds also cause inflammation and are generally used as artificial inflammation inducers for the
screening of anti-inflammatory agents (Fujiki and Sugimura, 1987). Tumor promoter-induced
inflammation can be distinguished from acute inflammation, which is exudative and accompanied
by fibroblast proliferation and granulation. TPA not only causes inflammation but influences
mammalian cell growth (Nakamura et al., 1995), suggesting that the molecular basis of the
inflammation is the pol reaction related to cell proliferation. Here, we investigated the mode of action
of stevioside, isosteviol and related compounds against pols and DNA metabolic enzymes, which are
responsible for DNA replication leading to cell proliferation and DNA repair/recombination and,
then, the relation between the degree of the cell growth inhibitory effect and the anti-inflammatory
activity.
We report here that one of the compounds, isosteviol (compound 3), is a potent inhibitor of
mammalian pols and human DNA topoisomerase II (topo II). Pols and topos were suggested to be
important cellular targets for chemical intervention in the development of anti-cancer agents. Isosteviol
inhibited both the growth of human cancer cells and the TPA-induced inflammation. Therefore,

Y. Mizushina et al. / Life Sciences 77 (2005) 2127–2140
2129
2
Glc Glc
OH
O
O
H
H
H
H
H
H
HOOC
HOOC
Glc OOC
1
2
3
OH
HO
O
O
O
H
H
H
OH
H
H
H
HOOC
HOOC
HOOC
4
5
6
CH2OH
OH
O
O
O
H
H
H
O
H
H
H
HOOC
MeOOC
HOOC
7
8
9
O
OH
OH
H
H
H
OH
H
H
H
MeOOC
MeOOC
MeOOC
10
11
12
Fig. 1. Chemical structure of isosteviol and related compounds. Compound 1: stevioside, compound 2: steviol, compound 3:
isosteviol, compound 4: 11h-hydroxyisosteviol, compound 5: 7h-hydroxyisosteviol, compound 6: 12h-hydroxyisosteviol,
compound 7: 17-hydroxyisosteviol, compound 8: 7-oxoisosteviol, compound 9: methyl ent-15h,16h-epoxy-13-hydroxykauran-
1
9-oate, compound 10: isosteviol methyl ester, compound 11: methyl ent-16h-hydroxybeyeran-19-oate, and compound 12:
methyl ent-7a, 16h-dihydroxybeyeran-19-oate.
information concerning the structural characteristics of stevioside-based compounds may provide
valuable insight for the design of new anti-cancer agents.
Materials and methods
Materials
3
[
sized template-primers such as poly(dA) and oligo(dT)12–18 were purchased from Amersham
H]-dTTP (2V-deoxythymidine 5V-triphosphate) (43 Ci/mmol), nucleotides, and chemically synthe-
Biosciences (Buckinghamshire, UK). Supercoiled pUC19 plasmid DNA and pBR322 DNA were

2
130
Y. Mizushina et al. / Life Sciences 77 (2005) 2127–2140
obtained from Takara (Tokyo, Japan). All other reagents were of analytical grade and were purchased
from Nacalai Tesque Inc. (Kyoto, Japan).
Chemicals
Sources of stevioside (compound 1), isosteviol (ent-16-ketobeyeran-19-oic acid, compound 3), 11h-
hydroxyisosteviol (compound 4), 7h-hydroxyisosteviol (compound 5), 12h-hydroxyisosteviol (com-
pound 6), 17-hydroxyisosteviol (compound 7), and 7-oxoisosteviol (compound 8) were described in our
recent paper (Akihisa et al., 2004). Thus, stevioside (compound 1) was kindly donated by Horiuchi
Foods, Co., Ltd. (Tokyo, Japan). Isosteviol (compound 3) was obtained by hydrolysis of stevioside
(compound 1) with dilute hydrochloric acid. Incubation of compound 3 with Aspergillus niger IFO 4414
yielded 7h-hydroxyisosteviol (compound 5), 11h-hydroxyisosteviol (compound 4), and 12h-hydro-
xyisosteviol (compound 6). Biotransformation of compound 3 by Glomerella cingulata IFO 9767
produced 17-hydroxyisosteviol (compound 7). 7-Oxoisosteviol (compound 8) was obtained from
biotransformation of compound 3 by Mortierella elongate IFO 8570.
Steviol (compound 2) was prepared from microbial transformation of stevioside (compound 1) by
using the fungus A. niger IFO 4414 (Subrahmanyam et al., 1999). Methyl ent-15h,16h-epoxy-13-
hydroxykauran-19-oate (compound 9) was obtained from compound 2 by epoxidation with m-
chloroperbenzoic acid in dichloromethane followed by esterification with diazomethane. Esterification
with diazomethane of isosteviol (compound 3) afforded isosteviol methyl ester (compound 10) (Avent et
al., 1990a,b) and reduction of compound 10 with NaBH in ethanol yielded methyl ent-16h-
4
hydroxybeyeran-19-oate (compound 11) (Ali et al., 1992). Reduction of 7h-hydroxyisosteviol (3) with
NaBH followed by esterification by diazomethane gave methyl ent-7a,16h-dihydroxybeyeran-19-oate
4
(compound 12) (Ali et al., 1992). Identification of compounds 2, 9, 10, 11, and 12 was performed by
mass spectral and H NMR spectral comparison with the corresponding compounds in the literature.
1
Enzymes
Pol a was purified from the calf thymus by immuno-affinity column chromatography as described
previously (Tamai et al., 1988). Recombinant rat pol h was purified from E. coli JMph5 as described by
Date et al. (1988). Recombinant His-pol E was overexpressed and purified according to methods
described previously (Shimazaki et al., 2002). Pol I (a-like) and II (h-like) from a higher plant,
cauliflower inflorescence, were purified according to the methods outlined by Sakaguchi et al. (1980).
The Klenow fragment of pol I from E. coli and human immunodeficiency virus (HIV) type-1 reverse
transcriptase (recombinant) was purchased from Worthington Biochemical Corp. (Freehold, NJ, USA).
T7 RNA polymerase was purchased from Takara (Kyoto, Japan). Purified human placenta topo I and II
(2 units/ml) were obtained from TopoGen, Inc. (Columbus, OH, USA). Bovine pancreas deoxyribo-
nuclease I was purchased from Stratagene Cloning Systems (LaJolla, CA, USA).
DNA polymerase assays
Activities of pols were measured using methods described previously (Mizushina et al., 1996, 1997).
The substrates of pols used were poly(dA)/oligo(dT)12–18 and dTTP as template-primer DNA and
nucleotide substrate, respectively. Isosteviol and related compounds were dissolved in dimethylsulfoxide

Y. Mizushina et al. / Life Sciences 77 (2005) 2127–2140
2131
(DMSO) at various concentrations and sonicated for 30 s. Aliquots of 4 Al of sonicated samples were
mixed with 16 Al of each enzyme (0.05 units) in 50 mM Tris–HCl (pH7.5) containing 1 mM
dithiothreitol, 50% glycerol and 0.1 mM EDTA and kept at 0 8C for 10 min. These inhibitor–enzyme
mixtures (8 Al) were added to 16 Al of each of the enzyme standard reaction mixtures and incubation was
carried out at 37 8C for 60 min. The activity without the inhibitor was considered 100% and the
remaining activity at each concentration of inhibitor was determined relative to this value. One unit of
pol activity was defined as the amount of enzyme that catalyzed the incorporation of 1 nmol of
deoxyribonucleotide triphosphates (i.e., dTTP) into synthetic template-primers (i.e. poly(dA)/
oligo(dT)12–18, A/T=2/1) in 60 min at 37 8C under normal reaction conditions for each enzyme
(Mizushina et al., 1996, 1997).
DNA topoisomerase assays
Relaxation activity of topo II was determined by detecting the conversion of supercoiled plasmid
DNA to its relaxed form (Spitzner et al., 1990). The topo II reaction was performed in 20 Al of a reaction
mixture containing 50 mM Tris–HCl buffer (pH 8.0), 120 mM KCl, 10 mM MgCl , 0.5 mM ATP, 0.5
2
mM dithiothreitol, supercoiled pUC19 DNA (0.25 Ag), 2 Al of inhibitor solution (10% DMSO) and 2
units of topo II. The reaction mixtures were incubated at 37 8C for 30 min and terminated by adding 2
Al of loading buffer consisting of 5% sarkosyl, 0.0025% bromophenol blue and 25% glycerol. The
mixtures were subjected to 1% agarose gel electrophoresis in TAE (Tris–acetate–EDTA) running buffer.
The agarose gel was stained with ethidium bromide (EtBr) and DNA was visualized on a UV
transilluminator. The relaxation activity of topo I was analyzed in the same manner except that the
reaction mixtures contained 10 mM Tris–HCl (pH 7.9), pBR322 DNA (0.25 Ag), 1 mM EDTA, 150 mM
NaCl, 0.1% bovine serum albumin (BSA), 0.1 mM spermidine, 5% glycerol and 2 units of topo I. One
unit was defined as the amount of enzyme capable of relaxing 0.25 Ag of DNA in 15 min at 37 8C.
Other enzyme assays
Activities of T7 RNA polymerase and bovine deoxyribonuclease I were measured in standard assays
according to the manufacturer’s specifications as described by Nakayama and Saneyoshi (1985) and Lu
and Sakaguchi (1991), respectively. Telomerase activity was determined using the polymerase chain
reaction (PCR)-based telomeric repeat amplification protocol as described (Kim et al., 1994) with some
modifications (Ueno et al., 2000).
Investigation of cytotoxicity in human cancer cells
To investigate the effects of isosteviol and related compounds in vivo, a human T cell acute
lymphoblastic leukemia cell line, MOLT-4, a human B cell acute lymphoblastoid leukemia cell line,
BALL-1, and a human gastric cancer cell line, NUGC-3, were obtained from Health Science Research
Bank (Osaka, Japan). The cells were routinely cultured in RPMI1640 medium supplemented with 10%
fetal calf serum, 100 Ag/ml streptomycin, 100 unit/ml penicillin, and 1.6 mg/ml NaHCO at 37 8C in a
3
humidified atmosphere of 5% CO –95% air. The cytotoxicity of the compounds was investigated as
follows. High concentrations (10 mM) of the compounds were dissolved in DMSO and stocked.
2
3
Approximately 5Â10 cells per well were inoculated in 96-well microplates, then the stock solution of

2
132
Y. Mizushina et al. / Life Sciences 77 (2005) 2127–2140
the compound was diluted to various concentrations and applied to each well. After incubation for 48 h,
the survival rate was determined by MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium
bromide) assay (Mosmann, 1983).
Anti-inflammatory assay
The mouse inflammatory test was performed according to Gschwendt’s method (Gschwendt et al.,
1
984). This experiment complied with the regulations concerning animal experimentation and the care
of experimental animals of Kobe-Gakuin University. Briefly, a methanol solution of the test compound
250 or 500 Ag/20 Al) was applied to the inner part of the mouse ear. Thirty minutes after the test
(
compound was applied, a TPA (12-O-tetradecanoylphorbol-13-acetate) solution (0.5 Ag/20 Al of
acetone) was applied to the same part of the ear. To the other ear of the same mouse, methanol and a
TPA solution were applied as a control. After 7 h, a disk (6 mm diameter) was obtained from the ear
and weighed. The inhibitory effect (IE) is presented as the ratio of the weight increase of the ear disks:
IE: {[(TPA only)À(tested compound plus TPA)]/[(TPA only)À(vehicle)]Â100}.
Computational analysis of isosteviol derivatives
A compound model was constructed and simple-minimized. Compound models were simulated with
force field parameters based on the consistent valence force field (CVFF). Group-based cutoffs, 0.95 nm
for van der Waals and 0.95 nm for Coulomb interactions, were introduced. The temperature was set at
2
98 K. Calculations based on simulation images were carried out using INSIGHT II (Accelrys Inc., San
Diego, CA, USA). Electrostatic potentials on the surface of compounds were analyzed with WebLab
ViewerLite (version 3.2, Accelrys Inc., San Diego, CA, USA) software.
Results and discussion
Effects of isosteviol and related compounds on DNA polymerases and other DNA metabolic enzymes
We, as a primary screening for inhibition of the DNA metabolic enzymes, tested the effects of
isosteviol and related compounds (compounds 1–12, Fig. 1) on mammalian pol a, h and E and
plant and prokaryotic pols. We decided that the compound, of which the IC50 values were lower
than 200 AM, was an inhibitor of the enzyme activity. As shown in Table 1, compound 3
(isosteviol) inhibited the activity of mammalian pols (a, h and E) but did not influence the
activities of higher plant (cauliflower) pol I (a-like polymerase) and II (h-like polymerase) and
prokaryotic pols such as the Klenow fragment of E. coli pol I. The other compounds, 1–2 and 4–
12, had no effect on the mammalian pols (Table 1). Human topo II activity was inhibited by
compounds 1, 3 and 7 (Table 1). On the other hand, human topo I activity was not influenced by
any of the compounds at all. The compounds had little effect on the activities of higher plant
(cauliflower) pol I (a-like polymerase) and II (h-like polymerase), prokaryotic pols such as the
Klenow fragment of E. coli pol I, and DNA metabolic enzymes such as T7 RNA polymerase,
human immunodeficiency virus (HIV) type-1 reverse transcriptase, human telomerase, and bovine
deoxyribonuclease I (Table 1).

Y. Mizushina et al. / Life Sciences 77 (2005) 2127–2140
2133
Table 1
IC50 values for isosteviol and related compounds on the activities of various DNA polymerases and other DNA metabolic
enzymes
Enzyme
Compound (AM)
1
2
3
4
5
6
7
8
9
10
11
12
Calf pol a
Rat pol h
Human pol E
Cauliflower pol I
N1000 N1000
N1000 N1000
N1000 N1000
64.0 N1000 N1000 N1000 N1000 N1000 N1000 N1000 N1000 N1000
177 N1000 N1000 N1000 N1000 N1000 N1000 N1000 N1000 N1000
103 N1000 N1000 N1000 N1000 N1000 N1000 N1000 N1000 N1000
a
N1000 N1000 N1000 N1000 N1000 N1000 N1000 N1000 N1000 N1000 N1000 N1000
N1000 N1000 N1000 N1000 N1000 N1000 N1000 N1000 N1000 N1000 N1000 N1000
N1000 N1000 N1000 N1000 N1000 N1000 N1000 N1000 N1000 N1000 N1000 N1000
b
Cauliflower pol II
E. coli pol I
T7 RNA polymerase N1000 N1000 N1000 N1000 N1000 N1000 N1000 N1000 N1000 N1000 N1000 N1000
c
HIV-RT
N1000 N1000 N1000 N1000 N1000 N1000 N1000 N1000 N1000 N1000 N1000 N1000
N1000 N1000 N1000 N1000 N1000 N1000 N1000 N1000 N1000 N1000 N1000 N1000
N1000 N1000 N1000 N1000 N1000 N1000 N1000 N1000 N1000 N1000 N1000 N1000
Human telomerase
Human topo I
Human topo II
90.0 N1000
190.0 N1000 N1000 N1000
82.5 N1000 N1000 N1000 N1000 N1000
d
Bovine DNase I
N1000 N1000 N1000 N1000 N1000 N1000 N1000 N1000 N1000 N1000 N1000 N1000
a
a-like DNA polymerase.
h-like DNA polymerase.
Reverse transcriptase.
Deoxyribonuclease I.
b
c
d
These observations suggested that some of the stevioside-based compounds could be inhibitors of
mammalian pols and human topo II and, of these twelve compounds, isosteviol (compound 3) was of
greatest interest as an inhibitor of the mammalian pols and topo II. Structurally, all compounds consist of a
tetracyclic diterpenoid with a beyerane skeleton and one carboxyl group (Fig. 1). Compound 1 is a
glycoside formed by three glucose molecules, compounds 2–8 have one free carboxyl group and
compounds 9–12 have one methyl ester of a carboxyl group. Compound 3 (isosteviol) has one ketone
group and no hydroxyl groups. From the results in Table 1, no hydroxyl groups, one carboxyl group and
one ketone group of diterpene acid might be considered to be important for the inhibition of mammalian
pols.
To determine whether a tetracyclic diterpenoid such as compound 3 binds to DNA, the
fluorescence emission spectra of ethidium bromide (EtBr) in the presence of calf thymus double-
stranded DNA (dsDNA) and the compound were measured. When EtBr was intercalated with
dsDNA, the fluorescence of the EtBr–dsDNA complex increased and the maximum emission
wavelength was 595 nm. At high concentrations (i.e., 1000 AM) of compound 3, no decrease in the
fluorescence of EtBr was observed (data not shown). Thus, compound 3 did not bind and
intercalate with the dsDNA, suggesting that it must inhibit the activity by interacting with the
enzyme directly.
Therefore, we concentrated on the properties of compound 3 in the latter part of this study.
Effects of compound 3 on the activities of mammalian DNA polymerases and human DNA
topoisomerases
Fig. 2 shows the inhibition dose–response curves of compound 3 against mammalian pols and
human topos. Pol a is a replicative pol and pol h and E are repair and recombination related pols.

2
134
Y. Mizushina et al. / Life Sciences 77 (2005) 2127–2140
1
00
80
60
40
20
0
0
50
100
150
200
Compound 3 (µM)
Fig. 2. Dose–response curves for the inhibition of mammalian DNA polymerases and human DNA topoisomerases by
compound 3. The enzymes used were calf pol a (closed circle), rat pol h (closed triangle), human pol E (closed square), human
topo I (open square) and human topo II (open circle). The pol and topo activities were measured as described in the Materials
and methods section. The enzyme activity in the absence of the compound was taken as 100%.
The compound was effective at inhibiting calf pol a, rat pol h, human pol E and human topo II
activities and the inhibition was dose-dependent, with 50% inhibition for pol a, h, E and topo II
observed at doses of 64.0, 177, 103 and 190 AM, respectively. The inhibitory activity by compound
3
was strongest toward pol a among the DNA metabolic enzymes tested (Fig. 2). Since the IC50
value of aphidicolin, which was an inhibitor of replicative pols such as pols a, y and q, for pol a
was 20–40 AM under the same conditions (Ikegami et al., 1978), compound 3 seemed not to be such
a strong inhibitor. On the other hand, the compound did not influence the activity of human topo I.
These observations indicated that compound 3 was a selective inhibitor of mammalian pols,
especially pol a.
Mode of DNA polymerase a inhibition by compound 3
To elucidate the inhibitory mechanism, the extent of inhibition was measured as a function of the
concentration of either the DNA template-primer or deoxyribonucleoside triphosphates (dNTP) in the
absence or presence of compound 3 (Fig. 3). In the kinetic analyses, poly(dA)/oligo(dT)12–18 and
dTTP were used as the DNA template-primer and nucleotide substrate, respectively. Double
reciprocal plots of the results indicated that the compound 3-mediated inhibition of the pol a activity
was non-competitive with both the DNA template-primer and nucleotide substrate. In the case of the
DNA template-primer, 35.7% and 62.5% decreases in maximum velocity (Vmax) were observed in the
presence of 25 and 50 AM compound 3, respectively, whereas the apparent Michaelis constant (Km)
was unchanged at 13.0 AM (Fig. 3A). The Km for the nucleotide substrate was unchanged at 1.65
AM and the Vmax for the nucleotide substrate decreased from 29.2 to 17.2 pmol/h in the presence of
5
0 AM of compound 3 (Fig. 3B). When activated DNA and dNTP were used as the DNA template-
primer and nucleotide substrate, respectively, the mode of inhibition of pol a by compound 3 was the

Y. Mizushina et al. / Life Sciences 77 (2005) 2127–2140
2135
(A)
(B)
0
0
0
.4
.3
.2
0
2
µ
M
0
2
5
µ
M
0
0
0
.4
.3
.2
5
µ
M
M
5
µ
M
M
50
µ
0
µ
0.1
0.1
-0.5
0
0.5
1
1.5 -1
0
1
2
3
-1
1 / S (µM-1) dTTP
1
/ S (µM ) poly(dA)/oligo(dT)
12-18
(C)
(D)
0
0
0
.4
.3
.2
0
0
.33
.4
µ
M dTTP
0
1
1
2
4
1
.67
M DNA
.33 M DNA
µ
M DNA
0
.4
.3
.2
µ
M dTTP
M dTTP
µ
0.5
µ
µ
0
0
0
1
1
1
.67
µ
M dTTP
µ
µ
M DNA
M DNA
µ
M dTTP
.25
.67
µ
µ
M dTTP
M dTTP
0 M DNA
µ
0.1
0.1
Ki
2.5 M dTTP
µ
Ki
-40
-20
0
20
40
-75
-50
-25
0
25
50
Compound 3 (µM)
Compound 3 (µM)
Fig. 3. Kinetic analyses of the inhibition of DNA polymerase a by compound 3. (A and B) Lineweaver–Burk double-reciprocal
plots obtained by varying DNA template-primer concentrations (A) and nucleotide substrate concentrations (B). Activity of
human pol a was assayed in the absence (square) or presence of 25 (circle) and 50 (triangle) AM of compound 3. (C and D) The
inhibition constant (K
i
) was determined as 27.5 and 57.0 AM from a Dixon plot made on the basis of the data for A and B,
respectively. The amount of human pol a in the assay mixture was 0.05 units.
same as when poly(dA)/oligo(dT)12–18 and dTTP were used (data not shown). These findings
suggested that the site where compound 3 binds to pol a is different from the DNA template-primer
and nucleotide substrate binding sites of pol a.
The inhibition constant (K ) for pol a, obtained from the Dixon plots in Fig. 3C and D, was found to
i
be 27.5 and 57.0 AM for the DNA template-primer and nucleotide substrate, respectively.
Effects of isosteviol and related compounds on cultured human cancer cells
The pols and topos have recently emerged as important cellular targets for chemical intervention in the
development of anti-cancer agents. The stevioside, isosteviol and related compounds could, therefore, be

2
136
Y. Mizushina et al. / Life Sciences 77 (2005) 2127–2140
useful in chemotherapy. We tested the cytotoxic effect of the twelve compounds against a human T cell
acute lymphoblastic leukemia cell line, MOLT-4, in vitro.
As shown in Fig. 4A, 200 AM of compound 3 had a potent growth inhibitory effect on this cancer cell
line, whereas the other compounds did not prevent cell growth. The MOLT-4 cell growth inhibition by
compound 3 was dose-dependent and the concentration required for the LD50 was 84 AM (Fig. 4B).
(A)
Compound 1
Compound 2
Compound 3
Compound 4
Compound 5
Compound 6
Compound 7
Compound 8
Compound 9
Compound 10
Compound 11
Compound 12
0
20
40
60
80
100
Relative activity (%)
(B)
1
00
8
6
4
2
0
0
0
0
0
0
50
100
150
200
Compound 3 (µM)
Fig. 4. Effects of isosteviol and related compounds on proliferation of human cancer cells. (A) Inhibition of cell growth of the
human T cell acute lymphoblastic leukemia cell line, MOLT-4, by twelve isosteviol and related compounds (200 AM each). (B)
Dose–response curves for the inhibition of human cancer cell growth by compound 3. MOLT-4 (circle), BALL-1 (human B cell
acute lymphoblastoid leukemia cell line, triangle) and NUGC-3 (human gastric cancer cell line, square). The assays were carried
out under the conditions described in Materials and methods with the indicated concentrations of compound 3. The rate of
survival was determined by MTT assay (Mosmann, 1983). Values are shown as the meanFSEM for three independent
experiments.

Y. Mizushina et al. / Life Sciences 77 (2005) 2127–2140
2137
Compound 3 also inhibited the cell growth of BALL-1 which is a human B cell acute lymphoblastoid
leukemia cell line and NUGC-3 which is a human gastric cancer cell line, with an LD50 of 87 and 167
AM, respectively. The dose–response curve for MOLT-4 was almost the same shape as that for BALL-1,
indicating that they do not selectively inhibit either B cells or T cells (Fig. 4B). The inhibitory effect for
the non-adherent cell lines (i.e., MOLT-4 and BALL-1 cells) was stronger than that for an adherent cell
line (i.e., NUGC-3 cells). The LD50 value of compound 3 for lymphocyte cells was similar to the IC50
value in vitro for pol a. The inhibition curves of the pol enzymes and cell growth showed parallel dose-
dependent reductions. These observations suggested that the cell growth inhibition occurred in a manner
dependent on the enzyme inhibition. Compound 3 was, therefore, suggested to inhibit the activity of
pols, especially replicative pol a, in the intact cell. Since compound 3 also arrested the cells at S-phase
(data not shown), the compound might distribute in nucleus and influences the progression of the cell
cycle.
Effect of anti-inflammatory activity by compound 3
Using the mouse inflammatory test, we examined the anti-inflammatory activity of compound 3. An
application of 12-O-tetradecanoylphorbol-13-acetate (TPA) (0.5 Ag) to the mouse ear induced edema,
the weight increase of an ear disk 7 h after the application being 241%. As shown in Table 2,
pretreatment with 250 and 500 Ag of compound 5 resulted in the strongest reduction in TPA-induced
inflammation among the compounds tested, with an inhibitory effect (IE) of 25.8% and 53.0%,
respectively. Because the IE of compound 3 (500 Ag) was more than 50%, this compound is suggested to
have an effect on the inflammation. TPA is a compound that promotes tumorigenesis (Hecker, 1978) and
is generally used as an artificial inducer of inflammation (Fujiki and Sugimura, 1987; Nakamura et al.,
1
995). TPA-induced inflammation can be distinguished from acute inflammation and is accompanied by
fibroblastic proliferation and granulation. These results suggested that the molecular basis of the so-
called promotion process in oncogenesis also contains a biochemical process which requires pols.
Three-dimensional structures of isosteviol and related compounds
To obtain information about the molecular basis for the differential inhibition exhibited by stevioside,
isosteviol and the ten related compounds, computational analyses were performed using molecular
simulation and surface analysis software (Fig. 5). The three-dimensional molecular structures of the
compounds are shown in Fig. 5(1–12). A comparison of the electrostatic potential of these compounds
revealed a remarkable difference in their overall disposition and rapport. The electrostatic potential at
Table 2
Anti-inflammatory activities of compound 3 in the mouse ear inflammation test
Compound
Inhibitory effect (%) (FS.E.)
50 Ag/ear
25.8 (F5.4)*
2
500 Ag/ear
3
53.0 (F4.8)*
S.E. is shown in parentheses. A sample (250 or 500 Ag) was applied to one ear and, after 30 min, TPA (0.5 Ag) was applied to
both ears of the mouse. The edema was evaluated after 7 h, the inhibitory effect being expressed as a percentage ratio of the
edema. Five mice were used for each experiment. *Significantly different, P b0.05, using Student’s t-test.

2
138
Y. Mizushina et al. / Life Sciences 77 (2005) 2127–2140
Fig. 5. Computer graphics of isosteviol and related compounds. Compound 1: stevioside, compound 2: steviol, compound 3:
isosteviol, compound 4: 11h-hydroxyisosteviol, compound 5: 7h-hydroxyisosteviol, compound 6: 12h-hydroxyisosteviol,
compound 7: 17-hydroxyisosteviol, compound 8: 7-oxoisosteviol, compound 9: methyl ent-15h,16h-epoxy-13-hydroxykauran-
1
9-oate, compound 10: isosteviol methyl ester, compound 11: methyl ent-16h-hydroxybeyeran-19-oate, and compound 12:
methyl ent-7a, 16h-dihydroxybeyeran-19-oate. Stick models of compounds 1–12 were built using the graphics program
INSIGHT II (Accelrys Inc., San Diego, CA, USA). Carbons, oxygens and hydrogens are indicated in gray, red and white,
respectively. Electostatic potentials over molecular surfaces were analyzed using WebLab ViewerLite software (version 3.2,
Accelrys Inc., San Diego, CA, USA). Blue areas are positively charged, red areas are negatively charged, and white areas are
neutral (see Materials and methods).
each point on a constant electronic density surface (approximating the van der Waals surface for each
arrangement) was represented graphically with red corresponding to the regions where the electrostatic
potential was most negative and blue corresponding to the most positive regions.
As shown in Fig. 5(3), isosteviol (compound 3) consisted of a hydrophobic region (upper surface) and
two negative electrostatic charges on the O atom in the carboxyl group (under-left) and ketone group
(under-right). Other compounds have another negative or positive charge. The following molecular
features might be important for the inhibition of pol activity, cytotoxicity and anti-inflammation; [1]

Y. Mizushina et al. / Life Sciences 77 (2005) 2127–2140
2139
tetracyclic diterpenoid with a beyerane skeleton (i.e., hydrophobic region), [2] one carboxyl group, [3]
one ketone group, and [4] no hydroxyl groups.
As described previously, sulfo-glycolipids, which inhibited the activity of mammalian pols equally
(Hanashima et al., 2000; Ohta et al., 2000), were as effective as anti-cancer agents in vivo (Sahara et
al., 1997, 2002). Compound 3 could, therefore, be considered a possible candidate for an anti-cancer
agent.
Acknowledgements
We are grateful for the donations of calf pol a, rat pol h, and human pol E by Dr. M. Takemura of Mie
University (Mie, Japan), Dr. A. Matsukage of Japan Women’s University (Tokyo, Japan), and Dr. N.
Shimazaki and Dr. O. Koiwai of Tokyo University of Science (Chiba, Japan), respectively. This work
was supported in part by a grant-in-aid from Kobe Gakuin University Joint Research (B) (Y. M. and H.
Y.) and bHigh-Tech Research CenterQ Project for Private Universities: matching fund subsidy from
MEXT (Ministry of Education, Culture, Sports, Science and Technology), 2000–2004 (Y. M. and H. Y.).
Y. M. acknowledges grants-in-aid from the Uehara Memorial Foundation (Japan), the Hyogo Science
and Technology Association (Japan), and The Japan Food Chemical Research Foundation (Japan), and
Grant-in-Aid 16710161 for Scientific Research, MEXT (Japan).
References
Akihisa, T., Hamasaki, Y., Tokuda, H., Ukiya, M., Kimura, Y., Nishino, H., 2004. Microbial transformation of isosteviol
and inhibitory effects on Epstein–Barr virus activation of the transformation products. Journal of Natural Products 67,
4
07–410.
Ali, M.S., Hanson, J.R., de Oliveira, B.H., 1992. The biotransformation of some ent-beyeran-19-oic acids by Gibberella
fujikuroi. Phytochemistry 31, 507–510.
Avent, A.G., Hanson, J.R., de Oliveira, B.H., 1990a. Hydrolysis of the diterpenoid glycoside, stevioside. Phytochemistry 29,
2
712–2715.
Avent, A.G., Hanson, J.R., Hitchcock, P.B., de Oliveira, B.H., 1990b. The influence of a 15-hydroxy group on the
rearrangement reactions of steviol and its 16,17-epoxide. Journal of Chemical Society Perkin Trans 1, 2661–2665.
Date, T., Yamaguchi, M., Hirose, F., Nishimoto, Y., Tanihara, K., Matsukage, A., 1988. Expression of active rat DNA
polymerase h in Escherichia coli. Biochemistry 27, 2983–2990.
Fujiki, H., Sugimura, T., 1987. Advances in Cancer Research, vol. 49. Academic Press, Inc., London, pp. 223–264.
Gschwendt, M., Kittstein, W., Furstenberger, G., Marks, F., 1984. The mouse ear edema: a quantitatively evaluable assay for
tumor promoting compounds and for inhibitors of tumor promotion. Cancer Letters 25, 177–185.
Hanashima, S., Mizushina, Y., Ohta, K., Yamazaki, T., Sugawara, F., Sakaguchi, K., 2000. Structure–activity relationship of a
novel group of mammalian DNA polymerase inhibitors, synthetic sulfoquinovosylacylglycerols. Japanese Journal of Cancer
Research 91, 1073–1083.
Hecker, E., 1978. Carcinogenesis, vol. 2. Raben Press, NY, pp. 11–48.
Ikegami, S., Taguchi, T., Ohashi, M., Oguro, M., Nagano, H., Mano, Y., 1978. Aphidicolin prevents mitotic cell division by
interfering with the activity of DNA polymerase-a. Nature 275, 458–460.
Kim, N.W., Piatyszek, M.A., Prowse, K.R., Haley, C.B., West, M.D., Ho, P.L.C., Coviello, G.M., Wright, W.E., Weinrich, S.L.,
Shay, J.W., 1994. Specific association of human telomerase activity with immortal cells and cancer. Science 266, 2011–2015.
Kinghorn, A.D., 2002. In: Kinghorn, A.D. (Ed.), Stevia, the Genus Stevia. Taylor and Francis, London, pp. 1–17.
Lu, B.C., Sakaguchi, K., 1991. An endo-exonuclease from meiotic tissues of the basidiomycete Coprinus cinereus: its
purification and characterization. Journal of Biological Chemistry 266, 21060–21066.

2
140
Y. Mizushina et al. / Life Sciences 77 (2005) 2127–2140
Mizushina, Y., Tanaka, N., Yagi, H., Kurosawa, T., Onoue, M., Seto, H., Horie, T., Aoyagi, N., Yamaoka, M., Matsukage, A.,
Yoshida, S., Sakaguchi, K., 1996. Fatty acids selectively inhibit eukaryotic DNA polymerase activities in vitro. Biochimica
et Biophysica Acta 1308, 256–262.
Mizushina, Y., Yoshida, S., Matsukage, A., Sakaguchi, K., 1997. The inhibitory action of fatty acids on DNA polymerase h.
Biochimica et Biophysica Acta 1336, 509–521.
Mizushina, Y., Miyazaki, S., Ohta, K., Hirota, M., Sakaguchi, K., 2000. Novel anti-inflammatory compounds from Myrsine
seguinii, terpeno-benzoic acids, are inhibitors of mammalian DNA polymerases. Biochimica et Biophysica Acta 1475, 1–4.
Mizushina, Y., Hirota, M., Murakami, C., Ishidoh, T., Kamisuki, T., Shimazaki, N., Takemura, M., Perpelescu, M., Suzuki, M.,
Yoshida, H., Sugawara, F., Koiwai, K., Sakaguchi, K., 2003. Some anti-chronic inflammatory compounds are DNA
polymerase E-specific inhibitors. Biochemical Pharmacology 66, 1935–1944.
Mosmann, T., 1983. Rapid colorimetric assay for cellular growth and survival: application to proliferation and cytotoxicity
assays. Journal of Immunology Methods 65, 55–63.
Murakami, C., Ishijima, K., Hirota, M., Sakaguchi, K., Yoshida, H., Mizushina, Y., 2002. Novel anti-inflammatory compounds
from Rubus sieboldii, triterpenoids, are inhibitors of mammalian DNA polymerases. Biochimica et Biophysica Acta 1596,
1
93–200.
Nakamura, Y., Murakami, A., Ohto, Y., Torikai, K., Tanaka, T., Ohigashi, H., 1995. Suppression of tumor promoter-induced
oxidative stress and inflammatory responses in mouse skin by a superoxide generation inhibitor 1V-acetoxychavicol acetate.
Cancer Research 58, 4832–4839.
Nakayama, C., Saneyoshi, M., 1985. Differential inhibitory effects of 5-substituted 1-h-d-xylofuranosyluracil 5V-triphosphates
and related nucleotides on DNA-dependent RNA polymerases I and II from the cherry salmon (Oncorhynchus masou).
Journal of Biochemistry Tokyo 97, 1385–1389.
Ohta, K., Hanashima, S., Mizushina, Y., Yamazaki, T., Saneyoshi, M., Sugawara, F., Sakaguchi, K., 2000. Studies on a novel
DNA polymerase inhibitor group, synthetic sulfoquinovosylacylglycerols: inhibitory action on cell proliferation. Mutat
Research 467, 139–152.
Sahara, H., Ishikawa, M., Takahashi, N., Ohtani, S., Sato, N., Gasa, S., Akino, T., Kikuchi, K., 1997. In vivo anti-tumor effect
of 3-sulphonoquinovosyl 1V-monoacylglyceride isolated from sea urchin (Strongylocentrotus intermedius) intestine. British
Journal of Cancer 75, 324–332.
Sahara, H., Hanashima, S., Yamazaki, T., Takahashi, S., Sugawara, F., Ohtani, S., Ishikawa, M., Mizushina, Y., Ohta, K.,
Shimozawa, K., Gasa, S., Jimbow, K., Sakaguchi, K., Sato, N., Takahashi, N., 2002. Anti-tumor effect of chemically
synthesized sulfolipids based on sea urchin’s natural sulfonoquinovosylmonoacylglycerols. Japanese Journal of Cancer
Research 93, 85–92.
Sakaguchi, K., Hotta, Y., Stern, H., 1980. Chromatin-associated DNA polymerase activity in meiotic cells of lily and mouse.
Cell Structure and Function 5, 323–334.
Shimazaki, N., Yoshida, K., Kobayashi, T., Toji, S., Tamai, T., Koiwai, O., 2002. Over-expression of human DNA polymerase
E in E. coli and characterization of the recombinant enzyme. Genes to Cells 7, 639–651.
Spitzner, J.R., Chung, I.K., Muller, M.T., 1990. Eukaryotic topoisomerase II preferentially cleaves alternating purine–
pyrimidine repeats. Nucleic Acids Research 18, 1–11.
Subrahmanyam, C., Rao, B.V., Ward, R.S., Hursthouse, M.B., Hibbs, D.E., 1999. Diterpenes from the marine mangrove
Bruguiera gymnorhiza. Phytochemistry 51, 83–90.
Tamai, K., Kojima, K., Hanaichi, T., Masaki, S., Suzuki, M., Umekawa, H., Yoshida, S., 1988. Structural study of
immunoaffinity-purified DNA polymerase a-DNA primase complex from calf thymus. Biochimica et Biophysica Acta 950,
2
63–273.
Ueno, T., Takahashi, H., Oda, M., Mizunuma, M., Yokoyama, A., Goto, Y., Mizushina, Y., Sakaguchi, K., Hayashi, H., 2000.
Inhibition of human telomerase by rubromycins: implication of spiroketal system of the compounds as an active moiety.
Biochemistry 39, 5995–6002.