- Email: [email protected]
Correlation between reduction potentials and inhibitory effects on Epstein–Barr virus activation by emodin derivatives
Cancer Letters 241 (2006) 263–267 www.elsevier.com/locate/canlet
Correlation between reduction potentials and inhibitory effects on Epstein–Barr virus activation by emodin derivatives Junko Koyama a,*, Munetaka Inoue a, Izumi Morita a, Norihiro Kobayashi a, Toshiyuki Osakai b, Hoyoku Nishino c, Harukuni Tokuda c a
Faculty of Pharmaceutical Sciences, Kobe Pharmaceutical University, Higashinada, Kobe 658-8558, Japan b Department of Chemistry, Faculty of Science, Kobe University, Nada, Kobe 657-8501, Japan c Department of Molecular Biochemistry, Kyoto Prefectural, University of Medicine, Kyoto 602-8566, Japan Received 24 March 2005; received in revised form 19 October 2005; accepted 19 October 2005
Abstract As a continuation of studies using natural and synthetic products as cancer chemopreventive agents, we examined the reduction– oxidation potentials of hydroxylated emodin derivatives prepared from emodin in phosphate buffer at pH 7.2 using cyclic voltammetry. A significant correlation was found between the reduction potentials and number of the hydroxyl groups and the inhibitory effects of the hydroxylated emodin derivatives on Epstein–Barr virus early antigen activation. The electronic properties, i.e. LUMO energy and atomic charges of carbon at the 9-position (C9) and oxygen at the 11-position (O11), may also be useful for estimating the inhibitory effect on EBV-EA activation. q 2006 Elsevier Ireland Ltd. All rights reserved. Keywords: Epstein–Barr virus; Emodin; Reduction potential; Cyclic voltammetry
1. Introduction Quinones (anthraquinones, naphthoquinones, and hetero-naphthoquinones) are important naturally occurring pigments widely distributed in nature, which possesses anti-cancer and anti-microbial activities. Previous reports have demonstrated in vitro anti-tumor promoting activity of mono-, di-, and poly-substituted anthraquinones by determining the inhibitory effects on Epstein–Barr virus early antigen (EBV-EA) activation induced by 12-O-tetradecanoylphorbol-13-acetate (TPA) in Raji cells [1–4]. In studies of the structure– activity relationships for drugs, standard redox potential * Corresponding author. Tel.: C81 78 441 7549; fax: C81 78 441 7550. E-mail address: [email protected] (J. Koyama).
is important for determining physiological activities [5]. We employed cyclic voltammetry to determine standard redox potentials and reduction potentials of 19 anthraquinones, six bianthraquinones, nine naphthoquinones, and 19 azaanthraquinones at the physiological pH of 7.2, and found significant correlations between the standard redox (or reduction) potentials and inhibitory effects (log IC50) on EBV-EA activation [6–10]. Here, we report the anti-tumor promoting effects of 11 hydroxylated emodin derivatives prepared from emodin on EBV-EA activation, and the structure– activity relations between the inhibitory effects (log IC50) and redox potentials. Furthermore, electronic properties of emodin derivatives were calculated by the PM3 method using the CAChe MOPAC program [11], which indicated that steric energy, total energy, LUMO energy, and the atomic charges at C9 and O11 may be
0304-3835/$ - see front matter q 2006 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.canlet.2005.10.043
264
J. Koyama et al. / Cancer Letters 241 (2006) 263–267
useful parameter for estimating the inhibitory effect on EBV-EA activation. 2. Material and methods 2.1. Reagents and materials Emodin (1) was isolated from the root bark of Cassia siamea [12]. 1-Acetylemodin (2), 1,8-diacetylemodin (3), and 1,3,8-triacetylemodin (4) were obtained by acetylation of 1. 1,3,8-Triacetylemodic acid (5) was prepared from 4 by oxidation with chromium (IV) oxide, and hydrolysis of 5 gave emodic acid (6). The synthesis of 6-hydroxymethyl-1,3,8trihydroxyanthraquinone (7) was performed according to the method of Hirose [13]. 2-Hydroxyemodin (8), 4-hydroxyemodin (9), 5-hydroxyemodin (10), and 7-hydroxyemodin (11) were prepared by Banks’ method [14] (Fig. 1). Tissue culture reagents TPA and n-butyric acid were obtained from Nacalai Tesque. The EBV-genome-carrying lymphoblastoid cells (Raji cells derived from Burkitt’s lymphoma) were cultured in RPMI 1640 medium (SIGMA, R8758), as described elsewhere [15]. Spontaneous activation of EBV-EA in our subline of Raji cells was less than 0.1%. 2.2. Procedure for EBV-EA activation Inhibition of EBV-EA activation was assayed using Raji cells (virus non-producer), cultivated in 10% FBS RPMI 1640 medium. The Raji indicator cells (1!106 ml) were incubated at 37 8C for 48 h in 1 ml of medium containing n-butyric acid (4 mM, inducer), 2 ml of TPA (20 ng/ml in DMSO), and a known amount of test compound dissolved in DMSO. Smears were made from the cell suspension. The activated cells were stained by high-titer EBV-EA positive sera from
nasopharyngeal carcinoma (NPC) patients and detected by a conventional indirect immunofluorescence technique [16]. In each assay, at least 500 cells were counted, and experiments were repeated three times. EBV-EA inhibitory activity of the test compound was expressed by comparison with that of the positive control (100%), which was conducted with n-butyric acid (4 mM) plus TPA (32 pmol). For the experiments, EA induction was approximately 40%; this value was taken as the positive control (100%). 2.3. Electrochemical measurements Cyclic voltammetry was performed by a conventional three-electrode system using a laboratory-constructed microcomputer-controlled system, in which a potentiostat (Hokuto Denko, HA-301) controlled the working electrode potential. A plastic-formed-carbon (PFC) electrode with a surface area of 0.071 cm2 (BAS, PFCE-3), an Ag/AgCl (saturated NaCl) electrode, and a platinum coil electrode were used as the working, reference, and counter electrodes, respectively. Before recording each voltammogram, pretreatment of the working electrode was conducted as described previously [6,7] . The test solution was 2:1 (v/v) phosphate buffer (pH 7.2)– ethanol solution containing 0.05 mM of an emodin derivative, which was degassed with prepurified N2 gas prior to the voltammetric measurements. The electrolytic cell was waterjacketed to maintain a temperature of 25G0.1 8C. 2.4. Correlation coefficient The correlations of electrochemical and electronic parameters upon EBV-EA activation of the anthraquinones were determined using Pearson’s correlation coefficient.
Fig. 1. Structures of emodin derivatives.
J. Koyama et al. / Cancer Letters 241 (2006) 263–267 Table 1 Inhibitory effects of emodin derivatives on EBV-EA activation Compound
1 2 3 4 5 6 7 8 9 10 11
log IC50a
% to control (% viability) Molar ratio (to 32 pmol TPA) 1000
500
100
10
1.6 (60) 7.1 (60) 9.5 (60) 11.3 (60) 10.6 (60) 1.0 (60) 0 (60) 0 (60) 0 (60) 0 (60) 0 (60)
25 51 49.9 32.7 31.4 24.3 22.0 23.0 22.7 21.4 21.1
52.1 73.8 75.8 78.5 76.3 51 50.8 49.3 50.0 49.1 48.2
92.1 100 100 100 100 91.1 90.6 90.9 90.9 89.6 89.3
2.511 2.692 2.702 2.665 2.65 2.498 2.485 2.484 2.486 2.471 2.465
a
The molar ratio of test compound to TPA giving 50% inhibition against a positive control (100%) was defined as IC50.
265
voltage scan rate was increased (e.g. up to 200 mV/s), the second sharp reduction peak, accompanied by a small anodic peak, became larger than the first reduction peak; therefore, this peak pair can be considered as an adsorption wave due to the redox reaction of the emodin derivative adsorbed at the electrode surface. In contrast, the first broad reduction peak was usually proportional to the square root of the scan rate, indicating that the peak was limited mainly by diffusion of the emodin derivative. With the exception of compound 9, the corresponding anodic peak was detectable, but small compared with the broad cathodic peak, probably because of the instability of the reduction product. Accordingly, the first cathodic peak potential at 20 mV/s (Epc-1), which reflects the ‘thermodynamic’ redox potentials, was employed to examine any connections with EBV-EA activation. In Fig. 3, the values of log IC50 vs reduction potential (Epc-1 in V) of compounds 1–11 (B) and 10 polysubstituted anthraquinones (!, except the Table 2 Epc, Epa values of emodin derivatives
Fig. 2. Cyclic voltammograms of compounds 1 and 9 at a PFC electrode in 2:1 (v/v) 0.1 M phosphate buffer (pH 7.2)–Et OH. Voltage scan rate: 20 mV/s. 1: -----, 9: ———
Compound
Epc-1 (V)
Epc-2 (V)
Epa (V)
1 2 3 4 5 6 7 8 9 10 11
K0.473 K0.437 K0.405 K0.391 K0.431 K0.373 K0.411 K0.379 K0.446 K0.411 K0.376
K0.632 K0.649 ND ND ND K0.571 K0.629 K0.536 ND K0.725 K0.725
K0.246 K0.275 K0.208 K0.491 K0.492 K0.214 K0.249 K0.333 ND K0.300 K0.319
3. Results and discussion
ND, not detected.
11 Emodin derivatives were tested for the ability to inhibit in vitro EBV-EA activation induced by TPA in Raji cells; results are summarized in Table 1. Acetylation of the phenolic hydroxyl group of emodin (1) and emodic acid (6) resulted in a decrease of potency (1/ 2–4, 6/5). Deacetylation of the 3-acetylemodin derivatives decreased activation (4,5/2,3). Oxidized emodin derivatives (6–11) possessed greater inhibitory ability compared to emodin (1). Cyclic voltammograms of 1 and 9, obtained at 20 mV/s, are shown in Fig. 2. As represented by the voltammogram for compound 1, emodin derivatives (except for compounds 3, 4, 5, and 9) produced two reduction peaks, one sharp and the other broad (cathodic and anodic peak potentials are summarized in Table 2 at 20 mV/s). However, when the
Fig. 3. Regression plot of log IC50 and the first reduction potential at pH 7.2 of anthraquinones with their EBV-EA activation. Plot of log IC50 against Epc-1.
266
J. Koyama et al. / Cancer Letters 241 (2006) 263–267
Table 3 Electronic properties of emodin derivatives Compound
HOMO (eV)
LUMO (eV)
HOMO– LUMO (eV)
Steric energy (kcal/mol)
Total energy (hartree)
Charge of C9
Charge of O11
1 2 3 4 5 6 7 8 9 10 11 ra
K9.433 K9.391 K9.592 K9.811 K8.731 K9.682 K9.436 K9.211 K9.043 K9.019 K9.181 0.643
K1.661 K1.473 K1.342 K1.222 K1.424 K1.948 K1.644 K1.697 K1.799 K1.801 K1.642 0.844
K7.772 K7.918 K8.250 K8.589 K8.307 K7.734 K7.792 K7.514 K7.244 K7.218 K7.539 0.827
K23.49 K19.47 K15.23 K10.58 K19.40 K33.18 K25.04 K28.39 K25.19 K22.62 K24.89 0.781
K151.25 K175.86 K200.48 K225.13 K247.13 K173.78 K163.42 K163.46 K163.48 K163.49 K163.47 0.730
0.440 0.420 0.391 0.394 0.397 0.436 0.439 0.439 0.437 0.436 0.438 0.902
K0.418 K0.360 K0.303 K0.296 K0.289 K0.406 K0.414 K0.416 K0.418 K0.416 K0.415 0.906
a
r, correlation coefficient with IC50.
bianthraquinones [10]) are plotted. The plot reveals no correlation between the values of log IC50 and Epc-1. However, good correlation of log IC50 and Epc-1 of compounds 1–5 and poly-substituted anthraquinones was obtained. The log IC50 is represented by a regression line log IC50 Z 3:718 C 2:541Epc-1 ðn Z 15; r Z 0:826Þ
(1)
where n and r are the number of test compounds and correlation coefficient, respectively. Note that the compounds deviated from this correlation were oxidized emodin derivatives which contain more hydroxyl group than emodin. The regression analysis confirms an apparent correlation between log IC50, the number of hydroxyl groups, and Epc-1 of the emodin derivatives (1–11), the regression equation being log IC50 Z 2:672K0:117Epc-1 K0:059m ðn Z 11; r Z 0:922Þ
Acknowledgements This study was supported in part by Grants-in-Aid from the Ministry of Education, Science, Sports and Culture, and the Ministry of Health and Welfare, Japan, and also from the National Cancer Institute (CA 17625), USA.
(2)
where m is the number of hydroxyl groups (Fig. 1), and for the 21 anthraquinones: log IC50 Z 3:161 C 0:934Epc-1 K0:069m ðn Z 21; r Z 0:832Þ
significant correlation with log IC50. Good correlations also existed between the charges of C9 and O11 with log IC50. In conclusion, the reduction potentials determined at the physiological pH (7.2) and the number of the hydroxyl groups correlated with inhibitory effects on EBV-EA activation of hydroxylated emodin derivatives. The atomic charges of C9 and O11 are also useful parameters for predicting inhibitory activity. These results allow the evaluation of anti-tumor promoting activity of quinone derivatives by measuring reduction potentials without in vitro screening.
(3)
We examined the correlation of log IC50 with the electronic properties of the emodin derivatives. Table 3 shows the electronic properties, including HOMO, LUMO, difference between the HOMO and LUMO energies, total energy, steric energy, and the charges of C9 and O11. Among these electronic properties, total and steric energies demonstrate
References [1] T. Konoshima, M. Kozuka, J. Koyama, T. Okatani, K. Tagahara, H. Tokuda, Studies on inhibitors of skin tumor promotion, VI. Inhibitory effects of quinones on Epstein–Barr virus activation, J. Nat. Prod. 52 (1989) 987–995. [2] G.J. Kapadia, V. Balasubramanian, H. Tokuda, T. Konoshima, M. Takasaki, J. Koyama, et al., Anti-tumor promoting effects of naphthoquinone derivatives on short term Epstein–Barr virus early antigen activation assay and in mouse skin carcinogenesis, Cancer Lett. 113 (1997) 47–53. [3] K. Tagahara, J. Koyama, T. Ogura, T. Konoshima, M. Kozuka, H. Tokuda, et al., Electronic properties and inhibitory effects on Epstein–Barr virus activation of mono- and di-substituted anthraquinones, Chem. Express 7 (1992) 557–560.
J. Koyama et al. / Cancer Letters 241 (2006) 263–267 [4] J. Koyama, I. Morita, K. Tagahara, M. Ogata, T. Mukainaka, H. Tokuda, H. Nishino, Inhibitory effects of anthraquinones and bianthraquinones on Epstein–Barr virus activation, Cancer Lett. 170 (2001) 15–18. [5] K. Kano, T. Konse, N. Nishimura, T. Kubota, Electrochemical properties of adriamycin absorbed on a mercury electrode surface, Bull. Chem. Soc. Jpn 57 (1984) 2383–2390. [6] J. Koyama, K. Tagahara, T. Osakai, Y. Tsujino, S. Tsurumi, H. Nishino, H. Tokuda, Inhibitory effects on Epstein-Barr virus activation of anthraquinones: correlation with redox potentials, Cancer Lett. 115 (1997) 179–183. [7] J. Koyama, I. Morita, K. Tagahara, T. Osakai, H. Hotta, M.X. Yang, et al., Correlation with redox potentials and inhibitory effects on Epstein–Barr virus activation of azaanthraquinones, Chem. Pharm. Bull. 49 (2001) 1214–1216. [8] J. Koyama, I. Morita, N. Kobayashi, T. Osakai, H. Hotta, J. Takayasu, et al., Correlation of redox potentials and inhibitory effects on Epstein–Barr virus activation of naphthoquinones, Cancer Lett. 201 (2003) 25–30. [9] J. Koyama, I. Morita, N. Kobayashi, T. Osakai, H. Hotta, J. Takayasu, et al., Correlation of redox potentials and inhibitory effects on Epstein–Barr virus activation of 2-azaanthraquinones, Cancer Lett. 212 (2004) 1–6.
267
[10] J. Koyama, I. Morita, N. Kobayashi, T. Osakai, H. Nishino, H. Tokuda, Correlation between reduction potentials and inhibitory effects on Epstein–Barr virus activation of poly-substituted anthraquinones, Cancer Lett. 225 (2005) 193–198. [11] J.J.P. Stewart, Optimization of parameters for semi-empirical methods. I. Method (CAChe scientific, Computer-aided chemistry from the Oxford Molecular Group), J. Comp. Chem. 10 (1989) 209–220. [12] J. Koyama, I. Morita, K. Tagahara, M. Aqil, Bianthra-quinones from Cassia siamea, Phytochemistry 56 (2001) 849–851. [13] Y. Hirose, Y. Suehiro, Y. Furukawa, T. Murakami, Chemische Studien Uber naturliche Anthrachinone. II. Synthese von Citreoroseine, Fallacinol und Fallacinal, Chem. Pharm. Bull. 30 (1982) 4186–4188. [14] H.J. Banks, D.W. Cameron, W.D. Raverty, Hydroxylation of anthraquinones in oleum and in sulfuric acid, Aust. J. Chem. 31 (1978) 2271–2282. [15] Y. Ito, S. Yanase, J. Fujita, T. Harayama, M. Takashima, H. Imanaka, A short-term in vitro assay for promoter substances using human lymphoblastoid cells latently infected with Epstein-Barr virus, Cancer Lett. 13 (1981) 29–37. [16] G. Henle, W. Henle, Immunofluorescence in cells derived from Brukitt’s lymphoma, J. Bacteriol. 91 (1966) 1248–1256.
Correlation between reduction potentials and inhibitory effects on Epstein–Barr virus activation by emodin derivatives Junko Koyama a,*, Munetaka Inoue a, Izumi Morita a, Norihiro Kobayashi a, Toshiyuki Osakai b, Hoyoku Nishino c, Harukuni Tokuda c a
Faculty of Pharmaceutical Sciences, Kobe Pharmaceutical University, Higashinada, Kobe 658-8558, Japan b Department of Chemistry, Faculty of Science, Kobe University, Nada, Kobe 657-8501, Japan c Department of Molecular Biochemistry, Kyoto Prefectural, University of Medicine, Kyoto 602-8566, Japan Received 24 March 2005; received in revised form 19 October 2005; accepted 19 October 2005
Abstract As a continuation of studies using natural and synthetic products as cancer chemopreventive agents, we examined the reduction– oxidation potentials of hydroxylated emodin derivatives prepared from emodin in phosphate buffer at pH 7.2 using cyclic voltammetry. A significant correlation was found between the reduction potentials and number of the hydroxyl groups and the inhibitory effects of the hydroxylated emodin derivatives on Epstein–Barr virus early antigen activation. The electronic properties, i.e. LUMO energy and atomic charges of carbon at the 9-position (C9) and oxygen at the 11-position (O11), may also be useful for estimating the inhibitory effect on EBV-EA activation. q 2006 Elsevier Ireland Ltd. All rights reserved. Keywords: Epstein–Barr virus; Emodin; Reduction potential; Cyclic voltammetry
1. Introduction Quinones (anthraquinones, naphthoquinones, and hetero-naphthoquinones) are important naturally occurring pigments widely distributed in nature, which possesses anti-cancer and anti-microbial activities. Previous reports have demonstrated in vitro anti-tumor promoting activity of mono-, di-, and poly-substituted anthraquinones by determining the inhibitory effects on Epstein–Barr virus early antigen (EBV-EA) activation induced by 12-O-tetradecanoylphorbol-13-acetate (TPA) in Raji cells [1–4]. In studies of the structure– activity relationships for drugs, standard redox potential * Corresponding author. Tel.: C81 78 441 7549; fax: C81 78 441 7550. E-mail address: [email protected] (J. Koyama).
is important for determining physiological activities [5]. We employed cyclic voltammetry to determine standard redox potentials and reduction potentials of 19 anthraquinones, six bianthraquinones, nine naphthoquinones, and 19 azaanthraquinones at the physiological pH of 7.2, and found significant correlations between the standard redox (or reduction) potentials and inhibitory effects (log IC50) on EBV-EA activation [6–10]. Here, we report the anti-tumor promoting effects of 11 hydroxylated emodin derivatives prepared from emodin on EBV-EA activation, and the structure– activity relations between the inhibitory effects (log IC50) and redox potentials. Furthermore, electronic properties of emodin derivatives were calculated by the PM3 method using the CAChe MOPAC program [11], which indicated that steric energy, total energy, LUMO energy, and the atomic charges at C9 and O11 may be
0304-3835/$ - see front matter q 2006 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.canlet.2005.10.043
264
J. Koyama et al. / Cancer Letters 241 (2006) 263–267
useful parameter for estimating the inhibitory effect on EBV-EA activation. 2. Material and methods 2.1. Reagents and materials Emodin (1) was isolated from the root bark of Cassia siamea [12]. 1-Acetylemodin (2), 1,8-diacetylemodin (3), and 1,3,8-triacetylemodin (4) were obtained by acetylation of 1. 1,3,8-Triacetylemodic acid (5) was prepared from 4 by oxidation with chromium (IV) oxide, and hydrolysis of 5 gave emodic acid (6). The synthesis of 6-hydroxymethyl-1,3,8trihydroxyanthraquinone (7) was performed according to the method of Hirose [13]. 2-Hydroxyemodin (8), 4-hydroxyemodin (9), 5-hydroxyemodin (10), and 7-hydroxyemodin (11) were prepared by Banks’ method [14] (Fig. 1). Tissue culture reagents TPA and n-butyric acid were obtained from Nacalai Tesque. The EBV-genome-carrying lymphoblastoid cells (Raji cells derived from Burkitt’s lymphoma) were cultured in RPMI 1640 medium (SIGMA, R8758), as described elsewhere [15]. Spontaneous activation of EBV-EA in our subline of Raji cells was less than 0.1%. 2.2. Procedure for EBV-EA activation Inhibition of EBV-EA activation was assayed using Raji cells (virus non-producer), cultivated in 10% FBS RPMI 1640 medium. The Raji indicator cells (1!106 ml) were incubated at 37 8C for 48 h in 1 ml of medium containing n-butyric acid (4 mM, inducer), 2 ml of TPA (20 ng/ml in DMSO), and a known amount of test compound dissolved in DMSO. Smears were made from the cell suspension. The activated cells were stained by high-titer EBV-EA positive sera from
nasopharyngeal carcinoma (NPC) patients and detected by a conventional indirect immunofluorescence technique [16]. In each assay, at least 500 cells were counted, and experiments were repeated three times. EBV-EA inhibitory activity of the test compound was expressed by comparison with that of the positive control (100%), which was conducted with n-butyric acid (4 mM) plus TPA (32 pmol). For the experiments, EA induction was approximately 40%; this value was taken as the positive control (100%). 2.3. Electrochemical measurements Cyclic voltammetry was performed by a conventional three-electrode system using a laboratory-constructed microcomputer-controlled system, in which a potentiostat (Hokuto Denko, HA-301) controlled the working electrode potential. A plastic-formed-carbon (PFC) electrode with a surface area of 0.071 cm2 (BAS, PFCE-3), an Ag/AgCl (saturated NaCl) electrode, and a platinum coil electrode were used as the working, reference, and counter electrodes, respectively. Before recording each voltammogram, pretreatment of the working electrode was conducted as described previously [6,7] . The test solution was 2:1 (v/v) phosphate buffer (pH 7.2)– ethanol solution containing 0.05 mM of an emodin derivative, which was degassed with prepurified N2 gas prior to the voltammetric measurements. The electrolytic cell was waterjacketed to maintain a temperature of 25G0.1 8C. 2.4. Correlation coefficient The correlations of electrochemical and electronic parameters upon EBV-EA activation of the anthraquinones were determined using Pearson’s correlation coefficient.
Fig. 1. Structures of emodin derivatives.
J. Koyama et al. / Cancer Letters 241 (2006) 263–267 Table 1 Inhibitory effects of emodin derivatives on EBV-EA activation Compound
1 2 3 4 5 6 7 8 9 10 11
log IC50a
% to control (% viability) Molar ratio (to 32 pmol TPA) 1000
500
100
10
1.6 (60) 7.1 (60) 9.5 (60) 11.3 (60) 10.6 (60) 1.0 (60) 0 (60) 0 (60) 0 (60) 0 (60) 0 (60)
25 51 49.9 32.7 31.4 24.3 22.0 23.0 22.7 21.4 21.1
52.1 73.8 75.8 78.5 76.3 51 50.8 49.3 50.0 49.1 48.2
92.1 100 100 100 100 91.1 90.6 90.9 90.9 89.6 89.3
2.511 2.692 2.702 2.665 2.65 2.498 2.485 2.484 2.486 2.471 2.465
a
The molar ratio of test compound to TPA giving 50% inhibition against a positive control (100%) was defined as IC50.
265
voltage scan rate was increased (e.g. up to 200 mV/s), the second sharp reduction peak, accompanied by a small anodic peak, became larger than the first reduction peak; therefore, this peak pair can be considered as an adsorption wave due to the redox reaction of the emodin derivative adsorbed at the electrode surface. In contrast, the first broad reduction peak was usually proportional to the square root of the scan rate, indicating that the peak was limited mainly by diffusion of the emodin derivative. With the exception of compound 9, the corresponding anodic peak was detectable, but small compared with the broad cathodic peak, probably because of the instability of the reduction product. Accordingly, the first cathodic peak potential at 20 mV/s (Epc-1), which reflects the ‘thermodynamic’ redox potentials, was employed to examine any connections with EBV-EA activation. In Fig. 3, the values of log IC50 vs reduction potential (Epc-1 in V) of compounds 1–11 (B) and 10 polysubstituted anthraquinones (!, except the Table 2 Epc, Epa values of emodin derivatives
Fig. 2. Cyclic voltammograms of compounds 1 and 9 at a PFC electrode in 2:1 (v/v) 0.1 M phosphate buffer (pH 7.2)–Et OH. Voltage scan rate: 20 mV/s. 1: -----, 9: ———
Compound
Epc-1 (V)
Epc-2 (V)
Epa (V)
1 2 3 4 5 6 7 8 9 10 11
K0.473 K0.437 K0.405 K0.391 K0.431 K0.373 K0.411 K0.379 K0.446 K0.411 K0.376
K0.632 K0.649 ND ND ND K0.571 K0.629 K0.536 ND K0.725 K0.725
K0.246 K0.275 K0.208 K0.491 K0.492 K0.214 K0.249 K0.333 ND K0.300 K0.319
3. Results and discussion
ND, not detected.
11 Emodin derivatives were tested for the ability to inhibit in vitro EBV-EA activation induced by TPA in Raji cells; results are summarized in Table 1. Acetylation of the phenolic hydroxyl group of emodin (1) and emodic acid (6) resulted in a decrease of potency (1/ 2–4, 6/5). Deacetylation of the 3-acetylemodin derivatives decreased activation (4,5/2,3). Oxidized emodin derivatives (6–11) possessed greater inhibitory ability compared to emodin (1). Cyclic voltammograms of 1 and 9, obtained at 20 mV/s, are shown in Fig. 2. As represented by the voltammogram for compound 1, emodin derivatives (except for compounds 3, 4, 5, and 9) produced two reduction peaks, one sharp and the other broad (cathodic and anodic peak potentials are summarized in Table 2 at 20 mV/s). However, when the
Fig. 3. Regression plot of log IC50 and the first reduction potential at pH 7.2 of anthraquinones with their EBV-EA activation. Plot of log IC50 against Epc-1.
266
J. Koyama et al. / Cancer Letters 241 (2006) 263–267
Table 3 Electronic properties of emodin derivatives Compound
HOMO (eV)
LUMO (eV)
HOMO– LUMO (eV)
Steric energy (kcal/mol)
Total energy (hartree)
Charge of C9
Charge of O11
1 2 3 4 5 6 7 8 9 10 11 ra
K9.433 K9.391 K9.592 K9.811 K8.731 K9.682 K9.436 K9.211 K9.043 K9.019 K9.181 0.643
K1.661 K1.473 K1.342 K1.222 K1.424 K1.948 K1.644 K1.697 K1.799 K1.801 K1.642 0.844
K7.772 K7.918 K8.250 K8.589 K8.307 K7.734 K7.792 K7.514 K7.244 K7.218 K7.539 0.827
K23.49 K19.47 K15.23 K10.58 K19.40 K33.18 K25.04 K28.39 K25.19 K22.62 K24.89 0.781
K151.25 K175.86 K200.48 K225.13 K247.13 K173.78 K163.42 K163.46 K163.48 K163.49 K163.47 0.730
0.440 0.420 0.391 0.394 0.397 0.436 0.439 0.439 0.437 0.436 0.438 0.902
K0.418 K0.360 K0.303 K0.296 K0.289 K0.406 K0.414 K0.416 K0.418 K0.416 K0.415 0.906
a
r, correlation coefficient with IC50.
bianthraquinones [10]) are plotted. The plot reveals no correlation between the values of log IC50 and Epc-1. However, good correlation of log IC50 and Epc-1 of compounds 1–5 and poly-substituted anthraquinones was obtained. The log IC50 is represented by a regression line log IC50 Z 3:718 C 2:541Epc-1 ðn Z 15; r Z 0:826Þ
(1)
where n and r are the number of test compounds and correlation coefficient, respectively. Note that the compounds deviated from this correlation were oxidized emodin derivatives which contain more hydroxyl group than emodin. The regression analysis confirms an apparent correlation between log IC50, the number of hydroxyl groups, and Epc-1 of the emodin derivatives (1–11), the regression equation being log IC50 Z 2:672K0:117Epc-1 K0:059m ðn Z 11; r Z 0:922Þ
Acknowledgements This study was supported in part by Grants-in-Aid from the Ministry of Education, Science, Sports and Culture, and the Ministry of Health and Welfare, Japan, and also from the National Cancer Institute (CA 17625), USA.
(2)
where m is the number of hydroxyl groups (Fig. 1), and for the 21 anthraquinones: log IC50 Z 3:161 C 0:934Epc-1 K0:069m ðn Z 21; r Z 0:832Þ
significant correlation with log IC50. Good correlations also existed between the charges of C9 and O11 with log IC50. In conclusion, the reduction potentials determined at the physiological pH (7.2) and the number of the hydroxyl groups correlated with inhibitory effects on EBV-EA activation of hydroxylated emodin derivatives. The atomic charges of C9 and O11 are also useful parameters for predicting inhibitory activity. These results allow the evaluation of anti-tumor promoting activity of quinone derivatives by measuring reduction potentials without in vitro screening.
(3)
We examined the correlation of log IC50 with the electronic properties of the emodin derivatives. Table 3 shows the electronic properties, including HOMO, LUMO, difference between the HOMO and LUMO energies, total energy, steric energy, and the charges of C9 and O11. Among these electronic properties, total and steric energies demonstrate
References [1] T. Konoshima, M. Kozuka, J. Koyama, T. Okatani, K. Tagahara, H. Tokuda, Studies on inhibitors of skin tumor promotion, VI. Inhibitory effects of quinones on Epstein–Barr virus activation, J. Nat. Prod. 52 (1989) 987–995. [2] G.J. Kapadia, V. Balasubramanian, H. Tokuda, T. Konoshima, M. Takasaki, J. Koyama, et al., Anti-tumor promoting effects of naphthoquinone derivatives on short term Epstein–Barr virus early antigen activation assay and in mouse skin carcinogenesis, Cancer Lett. 113 (1997) 47–53. [3] K. Tagahara, J. Koyama, T. Ogura, T. Konoshima, M. Kozuka, H. Tokuda, et al., Electronic properties and inhibitory effects on Epstein–Barr virus activation of mono- and di-substituted anthraquinones, Chem. Express 7 (1992) 557–560.
J. Koyama et al. / Cancer Letters 241 (2006) 263–267 [4] J. Koyama, I. Morita, K. Tagahara, M. Ogata, T. Mukainaka, H. Tokuda, H. Nishino, Inhibitory effects of anthraquinones and bianthraquinones on Epstein–Barr virus activation, Cancer Lett. 170 (2001) 15–18. [5] K. Kano, T. Konse, N. Nishimura, T. Kubota, Electrochemical properties of adriamycin absorbed on a mercury electrode surface, Bull. Chem. Soc. Jpn 57 (1984) 2383–2390. [6] J. Koyama, K. Tagahara, T. Osakai, Y. Tsujino, S. Tsurumi, H. Nishino, H. Tokuda, Inhibitory effects on Epstein-Barr virus activation of anthraquinones: correlation with redox potentials, Cancer Lett. 115 (1997) 179–183. [7] J. Koyama, I. Morita, K. Tagahara, T. Osakai, H. Hotta, M.X. Yang, et al., Correlation with redox potentials and inhibitory effects on Epstein–Barr virus activation of azaanthraquinones, Chem. Pharm. Bull. 49 (2001) 1214–1216. [8] J. Koyama, I. Morita, N. Kobayashi, T. Osakai, H. Hotta, J. Takayasu, et al., Correlation of redox potentials and inhibitory effects on Epstein–Barr virus activation of naphthoquinones, Cancer Lett. 201 (2003) 25–30. [9] J. Koyama, I. Morita, N. Kobayashi, T. Osakai, H. Hotta, J. Takayasu, et al., Correlation of redox potentials and inhibitory effects on Epstein–Barr virus activation of 2-azaanthraquinones, Cancer Lett. 212 (2004) 1–6.
267
[10] J. Koyama, I. Morita, N. Kobayashi, T. Osakai, H. Nishino, H. Tokuda, Correlation between reduction potentials and inhibitory effects on Epstein–Barr virus activation of poly-substituted anthraquinones, Cancer Lett. 225 (2005) 193–198. [11] J.J.P. Stewart, Optimization of parameters for semi-empirical methods. I. Method (CAChe scientific, Computer-aided chemistry from the Oxford Molecular Group), J. Comp. Chem. 10 (1989) 209–220. [12] J. Koyama, I. Morita, K. Tagahara, M. Aqil, Bianthra-quinones from Cassia siamea, Phytochemistry 56 (2001) 849–851. [13] Y. Hirose, Y. Suehiro, Y. Furukawa, T. Murakami, Chemische Studien Uber naturliche Anthrachinone. II. Synthese von Citreoroseine, Fallacinol und Fallacinal, Chem. Pharm. Bull. 30 (1982) 4186–4188. [14] H.J. Banks, D.W. Cameron, W.D. Raverty, Hydroxylation of anthraquinones in oleum and in sulfuric acid, Aust. J. Chem. 31 (1978) 2271–2282. [15] Y. Ito, S. Yanase, J. Fujita, T. Harayama, M. Takashima, H. Imanaka, A short-term in vitro assay for promoter substances using human lymphoblastoid cells latently infected with Epstein-Barr virus, Cancer Lett. 13 (1981) 29–37. [16] G. Henle, W. Henle, Immunofluorescence in cells derived from Brukitt’s lymphoma, J. Bacteriol. 91 (1966) 1248–1256.