Planta 2008-04-0368 Neu-791-PDF1, 18.7.08/Druckerei Sommer
Cytotoxic Activity of C-Geranyl Compounds from Paulownia tomentosa Fruits Karel Ï Smejkal1, Petr Babula1, Tereza Ï Slapetovµ1, ÏemlicÏka1, Eleonora Brognara2, Stefano Dall'Acqua3, Milan Z Gabbriella Innocenti3, Josef CvacÏka4 1 Department of Natural Drugs, University of Veterinary and Pharmaceutical Sciences Brno, Brno, Czech Republic 2 Department of Biochemistry and Molecular Biology, University of Ferrara, Ferrara, Italy 3 Department of Pharmaceutical Sciences, University of Padua, Padua, Italy 4 Mass Spectrometry Group, Institute of Organic Chemistry and Biochemistry, v.v.i., Academy of Sciences of the Czech Republic, Prague, Czech Republic
Abstract !
The newly discovered 5,7-dihydroxy-6-geranylchromone (1) was isolated from Paulownia tomentosa fruit and subsequently characterized. The structure of the isolated compound was elucidated on the basis of extensive NMR experiments including HMQC, HMBC, COSY, and NOESY, as well as HR-MS, IR, and UV. The cytotoxicity of 1 was evaluated using a plant cell model represented by tobacco BY-2 cells. The other phytoconstituents (2 ± 8) previously isolated from P. tomentosa were similarly evaluated together with the known flavanones 10 and 11. The cytotoxicity (human erythro-leukaemia cell line K562) and activity on erythroid differentiation of compounds 2 ± 9 and 12 and 13 have also been evaluated. Acteoside (2) was determined to be the most toxic of the compounds tested on BY-2 cells, diplacone (6) on the K562 cell line. Some aspects of the relationship between the flavanone skeleton substitution and the metabolic activation necessary for a toxic effect are discussed.
Key words !
Paulownia tomentosa ´ Scrophulariaceae ´ Cytotoxicity ´ Flavanone ´ K562 ´ BY-2 Supporting information available online at http://www.thieme-connect.de/ejournals/toc/plantamedica
In our ongoing study of biologically active natural polyphenolic compounds, we have investigated the fruit of Paulownia tomentosa. Publications have reported mostly polyphenolic substances extracted from different parts of P. tomentosa in MeOH, EtOH and n-BuOH [1], [2], [3]. We report here the isolation and structural elucidation of the novel compound 5,7-dihydroxy-6geranylchromone (1). We also describe the determination of cytotoxic activities of P. tomentosa phenols by using a cell suspension culture of tobacco BY-2 cells. The activity against the human erythro-leukaemia cell line K562 and activity on erythroid differentiation have also been tested. An EtOH extract of P. tomentosa fruit was subjected to liquid-liquid fractionation and the fractions were repeatedly separated by CC on silica gel and further by preparative RP-HPLC. This exten-
Letter sive separation process resulted in the isolation of a new natural compound, the 5,7-dihydroxy-6-geranylchromone (1). The isolation of acteoside (2), isoacteoside (3) and the C-geranylflavonoids 4 ± 9, 12 and 13 from P. tomentosa has been described ear" lier [1], [2], [3], [4].l n The intensive bands at 2972 ± 2854 cm±1 in the IR spectrum of 1 indicated the presence of an aliphatic side chain in the structure of 1. Together with absorption bands at 1643 cm±1 (attributed to the CO stretch), 1617 ± 1448 cm±1 (the substituted aromatic system), 1374 cm±1 (the umbrella vibration of CH3), and bands in the region 1302 ± 1167 cm±1 (the aryl-alkyl ether bridge) these facts suggest a prenylated flavonoid [1], [2]. However the UV spectra showed unusual behavior with maxima at l = 210, 230 sh, 260, and 298 nm, which are not typical for a prenylated flavonoid. The HR-MS of compound 1 showed a protonated molecular ion [M + H]+ at m/z = 315.1602, indicating the molecular formula " Table 1) had a C19H22O4. The 1H-NMR spectrum of 1 in CDCl3 (l doublet at d = 7.73 (J = 6.05 Hz), a singlet at d = 6.37, and a further doublet at d = 6.23 (J = 6.05 Hz), integrating as one proton each. Diagnostic HMBC correlations were observed for the doublet at d = 7.73, with carbon resonances at d = 182.6 (C-4), 112.2 (C-3), and 157.1 (C-8a), and for the doublet at d = 6.23 with signals at d = 155.7 (C-2) and 106.4 (C-4a). Further correlations were observed between the proton signal at d = 6.37 (H-8) and carbon signals at d = 163.7 (C-7), 157.1 (C-8a), 110.5 (C-6) and 106.4 (C4a). These data support the presence of a chromone ring. Furthermore the 1H-NMR spectrum clearly revealed the presence of two singlets, with signals at d = 12.98 and 6.25 suggesting the presence of two phenolic OH groups (one chelated OH group at C-5; one at C-7). In addition, a set of signals characteristic for a geranyl side-chain was observed, namely two olefinic protons at d = 5.27 and 5.09 (m, one proton each), a benzylic proton signal at d = 3.44 (d, J = 7.38 Hz, two protons), two methylene groups at d = 2.09 ± 2.06 and 2.00 ± 1.97 (m, two protons each) and three methyl groups signals, at d = 1.83, 1.68 and 1.58 (s, three protons each). COSY, NOESY, and HMBC correlations confirmed the presence of the geranyl residue. HMBC correlations between the signal at d = 3.44 (C-1¢) with the carbons at d = 159.7 ppm (C-5), 163.7 ppm (C-7), and 110.5 ppm (C-6) revealed that the chromone was linked at C-6. On the basis of these data, compound 1 was characterized as 5,7-dihydroxy-6-geranylchromone. These data demonstrate that compound 1 has been obtained from a natural source for the first time. Compounds 2 and 3 were identified by comparing MS, UV, and IR absorption spectra and 1H- and 13C-NMR spectra with those in the literature [4]. Taxifolin (10) (Fluka, purity ³ 85 %) and 5,7,4¢trihydroxyflavanone (11) (Aldrich, purity > 99 %) used for the cytotoxicity assay were commercially available. A plant cell suspension culture of Nicotiana tabacum cv. Bright Yellow (BY-2) and of the K562 cell line were chosen as a suitable model for cytotoxicity testing. As shown in Fig. 1S (Supporting Information), exposure of tobacco BY-2 cells to the compounds being tested for 12, 24, and 48 hours produced a dose-dependent decrease in rate of growth and reduced cell viability in comparison with control cells. All of the compounds tested demonstrated a certain degree of toxicity depending on the applied dose; our results correspond partially with previously published studies of flavanones using various animal and tumor cell lines as model systems. To effectively compare the toxic effects of selected flavonoid compounds against the BY-2 cell line, the LD50 values were calculated at 12 and 48 hours after treatment. These LD50 values Letter ¼ Planta Med 2008; 74: 1 ± 4
1
2
Letter
Planta 2008-04-0368 Neu-791-PDF1, 18.7.08/Druckerei Sommer
" Table 2. Acteoside (2) showed the highest cyare presented in l totoxic effect. LD50 values of 52 mM and 42 mM were determined at 12 and 48 h, respectively. The cytotoxic activity of 2 is well known [5], [6]. The average toxicity of geranylchromone 1, which lacks the flavonoid phenyl ring B in its structure, confirms the importance of the flavonoid skeleton for cytotoxic activity [7]. A comparative study of flavones and flavanones has confirmed the
Table 1 1H- and 13C-NMR chemical shifts (d in ppm), indirect spin-spin coupling constants (J in Hz) and HMBC of 5,7-dihydroxy-6-geranylchromone (1) in CDCl3 at 303 K Position
dH, (J in Hz)
dC
HMBC
2 3 4 4a 5 6 7 8 8a 1¢ 2¢ 3¢ 4¢ 5¢ 6¢ 7¢ 8¢ 9¢ 10¢ OH-5 OH-7
7.73, d (6.1) 6.23, d (6.1)
155.7 112.2 182.6 106.4 159.7 110.5 163.7 95.4 157.1 20.2 120.9 140.8 39.6 25.6 123.6 132.0 25.7 16.1 18.4
3, 4, 9 2, 10
6.37, s 3.44, d (7.4) 5.27, m 1.97 ± 2.00, m 2.06 ± 2.09, m 5.09, m 1.68, s 1.83, s 1.58, s 12.98, s 6.25, s
Letter ¼ Planta Med 2008; 74: 1 ± 4
6, 1¢ 6, 8, 9, 1¢ 7, 9, 6,10 5, 6, 7, 2¢ 1¢, 3¢, 4¢, 9¢ 2¢, 3¢, 5¢, 6¢ 3¢, 4¢, 6¢, 7¢ 4¢, 5¢, 8¢, 10¢ 6¢, 7¢, 10¢ 2¢, 3¢, 4¢ 6¢, 7¢, 8¢
significance of the 2/3 double bond for cytotoxic activity [8]. After treatment for 12 h, taxifolin (9) was determined to be the most toxic compound of the flavonoids tested, followed by diplacone (6). These two compounds are flavonoids with ortho-dihydroxy substitution on the B ring. The para-hydroxy substituted B ring of mimulone (5), along with 3¢-O-methyldiplacone (7) and 3¢-O-methyldiplacol (4), are less toxic. The descending series of cytotoxicity concludes with 3¢-O-methyl-5¢-O-methyldiplacone (8) and 5,7,4¢-trihydroxyflavanone (11). When these cytotoxicity levels were compared with the values obtained after treatment for 48 hours; the sequence of compounds was significantly different. Taxifolin (10) remained the most toxic, but 3¢-Omethyl-5¢-O-methyldiplacone (8) demonstrated the same ability to inhibit growth and viability of cells. The toxicities of 3¢-methoxy-substituted compounds also increased and were slightly greater than those of diplacone (6), mimulone (5) and, 5,7,4¢-trihydroxyflavanone (11). The cytotoxic activity of flavonoid compounds depends primarily on the methylation and hydroxylation patterns of the skeleton and the type of the side chain [9], [10], [11], [12]. The most potent was found to be 2¢-hydroxy-substitution of the flavanone skeleton; different colorectal carcinoma cell lines were used for this cytotoxicity assay [13]. Comparison of the toxicities of flavanones 4 ± 8, 10 and 11 showed some structure-activity relationships. We can assume that a certain level of metabolic activation of the methoxy group is necessary for the toxic action, as it is clearly visible for 4, 7, and 8. Previous studies have described the metabolism of hydroxy and methoxy derivatives of flavonoids. Demethylation at position 4¢, and hydroxylation at position 3¢ occur when some dietary flavonoids are metabolized by different P-450 isoforms [14]. The metabolism of the flavanone glycoside naringin in the in vitro mixedculture model of human colonic microflora has been studied, and fission of ring C with subsequent degradation has been ob-
Letter
Planta 2008-04-0368 Neu-791-PDF1, 18.7.08/Druckerei Sommer
Table 2
Cytotoxicity of the compounds tested after 12 and 48 hours
Compound 5,7-Dihydroxy-6-C-geranylchromone (1) Acteoside (2) Isoacteoside (3) 3¢-O-Methyldiplacol (4) Mimulone (5) Diplacone (6) 3¢-O-methyldiplacone (7) 3¢-O-methyl-5¢-O-methyldiplacone (8) 3¢-O-methyl-5¢-O-hydroxydiplacone (9) Taxifolin (10) 5,7,4¢-trihydroxyflavanone (11) Tomentodiplacone (12) Tomentodiplacone B (13) Cisplatinum Ara-C a,b c
LD50 after
LD50 after
LD50 after
% of
12 hours [mM]a
48 hours [mM]b
72 hours [mM]c
differentiation
97 4 52 2 ±d 1066 89 4 82 4 102 6 107 5 ±d 63 3 114 6 ±d ±d 47 2 ±d
102 5 42 2 ±d 894 93 4 9190.5 4 80 3 62 3 ±d 62 3 106 5 ±d ±d 17 1 ±d
±d 30 19 29 10 21 13 13 8 4.4 6 16 5 24 5 64 ±d ± 33 10 27 15 ±d 0.3 0.1
±d 0 0 0 0 0 0 5 0 ±d ± 5 ± 10 7 ±d 60 ± 70e
As determined by the effect on the growth and the viability of tobacco BY-2 cells.
K562 cells, and erythroid differentiating activity at 10 mM.
d
Not determined.
e
Effect at 1 mM.
S.D. determined from three independent measurements.
served [15]. Metabolism of flavanones has also been observed in vivo in pigs [16]. From the increases in the cytotoxicity shown by compounds 8, 4 and 7 after 48 hours of treatment, we can deduce that the methoxy-substituted flavanones are probably metabolized by means of demethylation by the BY-2 cells. The similar LD50 values of compounds 5 and 6 after 48 hours indicate metabolism by hydroxylation at the 3 position. Comparing the results for compounds 5 and 11 shows that the presence of lipo philic substituent geranyl at position 6 reduces the cytotoxic activity of the otherwise similar compounds. The presence of the 3-OH substituent has almost no impact on the cytotoxic activity (compare compounds 4 and 7) after 12 hours, but the difference is evident after 48 hours of treatment; the 3-OH reduced the cytotoxic activity. The human erythro-leukaemia K562 cells have been treated with the compounds 2 ± 8 and 9 ± 11 at different concentrations. The cell viability has been evaluated after 3, 4, and 5 days of incubation. IC50 values were calculated to compare the cytotoxic " Table 2). Different levels of effects of the compounds tested (l cytotoxicity have been found. The geranylflavanones tested against K562 (4 ± 8, 9, 11) were found to be more cytotoxic than phenylpropanoids 2 and 3. Diplacon (6) was the most cytotoxic compound with an IC50 value of 4.4 mM. In the K562, the orthodihydroxy substitution of the flavanone B ring is most important for cytotoxic activity (6 and 8), para-hydroxy (5) and methoxy (4, 7, 9, 12, 13) substitution decrease the cytotoxic effect. Erythroid differentiation activity of compounds 2 ± 9, 11 and 12 has been established. Compounds 8, 11 and 12 showed weak activity at 10 mM, other compounds did not affect the erythroid differentiation or the effect was very low. Both active compounds 11 and 12 have modified geranyl side chain, 8 has a 3,5-dimethoxy-substituted B ring.
Materials and Methods
Plant material: The fruit of Paulownia tomentosa Steud. (Scrophulariaceae) was collected in the area of the University of Veterinary and Pharmaceutical Sciences Brno, Czech Republic, during October 2004. The plant was identified by Mr. Petr Babula, Department of Natural Drugs, Faculty of Pharmacy, University of Veterinary and Pharmaceutical Sciences Brno, Brno, Czech Republic. A voucher specimen (PT-04O) was deposited at the herbarium of the Department of Natural Drugs. Extraction and isolation: The extraction of P. tomentosa fruits and basic chromatographic separation of MeOH portion on silica were described previously [2]. Fractions G and J were combined and separated on silica with the help of flash chromatography (the mobile phase was C6H6:CHCl3:MeOH = 10 : 85 : 5; fractions of about 150 mL were collected). Fractions were combined to form G+J/10 ± 12 which was purified by semipreparative HPLC to give compound 1 (25 mg). For the mobile phase the solvent consisted of MeCN (A), and 0.2 % HCOOH (B), with gradient elution starting from A 50 % and B 50 %, and changing to A 100 % over 20 min; the flow rate was 5 mL.min±1. UV detection at l = 254 nm was used; fractions were collected according to the detector response. The isolation of the other compounds has been described previously [2]. The purity of compounds ranged between 96 ± 99 %, as determined from HPLC analysis. 5,7-Dihydroxy-6-geranylchromone (1): white powder (CHCl3, MeOH); m. p. 162 ± 166 C; UV: lmax (log e) = 260 (4.13), 296 (3.98) nm; IR (ATR): nmax = 3075, 2972, 2917, 2854, 1643, 1617, 1554, 1448, 1410, 1374, 1302, 1253, 1226, 1167, 1093, 1026 cm±1; " Table 1; HR-MS (ESI-TOF, posifor 1H- and 13C-NMR data, see l tive mode): m/z = 315.1606 (calcd. for C19H23O4 [M + H]+: 315.1596). For cytotoxicity assay on BY-2 cells, in vitro anti-proliferative assay on erythro-leukaemia K562 cells and in vitro induction of erythroid differentiation, see the Supporting Information.
!
The general experimental procedure has been described previously [1], [2]. For details please refer to the Supporting Information. Letter ¼ Planta Med 2008; 74: 1 ± 4
3
4
Letter
Planta 2008-04-0368 Neu-791-PDF1, 18.7.08/Druckerei Sommer
Acknowledgements !
Financial support of this work by the IGA VFU (grant No. 22/2006 to KS) and by the Ministry of Health (1A8666 to MD) is gratefully acknowledged.
References 1 Smejkal K, Grycova L, Marek R, Lemire F, Jankovska D, Forejtníkova H et al. C-Geranyl compounds from Paulownia tomentosa fruits. J Nat Prod 2007; 70: 1244 ± 8 2 Smejkal K, Chudík S, Kloucek P, Marek R, Cvacka J, Urbanova M et al. Antibacterial C-geranylflavonoids from Paulownia tomentosa Fruits. J Nat Prod; 2008: DOI: 10.1021/np070446 u 3 Asai T, Hara N, Kobayashi S, Kohshima S, Fujimoto Y. Geranylated flavanones from the secretion on the surface of the immature fruits of Paulownia tomentosa. Phytochemistry 2008; 69: 1234 ± 41 4 Kang KH, Jang SK, Kim BK, Park MK. Antibacterial phenylpropanoid glycosides from Paulownia tomentosa Steud. Arch Pharm Res 1994; 17: 470 ± 5 5 Lee KW, Kim HJ, Lee YS, Park HJ, Choi JW, Ha J et al. Acteoside inhibits human promyelocytic HL-60 leukemia cell proliferation via inducing cell cycle arrest at G(0)/G(1) phase and differentiation into monocyte. Carcinogenesis 2007; 28: 1928 ± 36 6 Pettit GR, Numata A, Takemura T, Ode RH, Narula AS, Schmidt JM et al. Antineoplastic agents 107. Isolation of acteoside and isoacteoside from Castilleja linariaefolia. J Nat Prod 1990; 53: 456 ± 8 7 Li NG, You QD, Huang XF, Wang JX, Guo QL, Chen XG et al. Synthesis and antitumor activity of small compounds structurally related to gambogic acid. Chin Chem Lett 2007; 18: 659 ± 62 8 Rao YK, Geethangili M, Fang S-H, Tzeng Y-M. Antioxidant and cytotoxic activities of naturally occurring phenolic and related compounds: A comparative study. Food Chem Toxicol 2007; 45: 1770 ± 6 9 Ko WG, Kang TH, Kim NY, Lee SJ, Kim YC, Ko GI et al. Lavandulylflavonoids: a new class of in vitro apoptogenic agents from Sophora flavescens. Toxicol In Vitro 2000; 14: 429 ± 33 10 Moridani MY, Galati G, O'Brien PJ. Comparative quantitative structure toxicity relationships for flavonoids evaluated in isolated rat hepatocytes and HeLa tumor cells. Chem Biol Interact 2002; 139: 251 ± 64 11 Hayashi K, Nakanishi Y, Bastow KF, Cragg G, Nozaki H, Lee K-H. Antitumor agents. 221. Buceracidins A and B, two new flavanones from Bucida buceras. J Nat Prod 2003; 66: 125 ± 7 12 Wätjen W, Weber N, Lou Y, Wang Z, Chovolou Y, Kampkötter A et al. Prenylation enhances cytotoxicity of apigenin and liquiritigenin in rat H4IIE hepatoma and C6 glioma cells. Food Chem Toxicol 2007; 45: 119 ± 24 13 Shen S-C, Ko CH, Tseng S-W, Tsai S-H, Chen Y-C. Structurally related antitumor effects of flavanones in vitro and in vivo: involvement of caspase 3 activation, p21 gene expression, and reactive oxygen species production. Toxicol Appl Pharmacol 2004; 197: 84 ± 5 14 Breinholt VM, Offord EA, Brouwer C, Nielsen SE, Brùsen K, Friedberg T. In vitro investigation of cytochrome P450-mediated metabolism of dietary flavonoids. Food Chem Toxicol 2002; 40: 609 ± 16 15 Rechner AR, Smith MA, Kuhnle G, Gibson GR, Debnam ES, Srai SKS et al. Colonic metabolism of dietary polyphenols: Influence of structure on microbial fermentation products. Free Radic Biol Med 2004; 36: 212 ± 25 16 Silberberg M, Gil-Izquierdo A, Combaret L, Remesy C, Scalbert A, Morand C. Flavanone metabolism in healthy and tumor-bearing rats. Biomed Pharmacother 2006; 60: 529 ± 35 17 Murashige T, Skoog F. A revised medium for rapid growth and bioassys with tobacco tissue cultures. Physiol Plant 1962; 15: 473 ± 97 18 Nagata T, Nemoto Y, Hasetawa SA. Tobacco BY-2 cell line as the Helacell in the cell biology of higher plants. Int Rev Cytol Surv Cell Biol 1992; 132: 1 ± 30 19 Bianchi N, Chiarabelli C, Borgatti M, Mischiatti C, Bergamini P, Fibach E et al. Accumulation of gamma-globin mRNA and induction of erythroid differentiation after treatment of human leukemia K562 cells with tallimustine. Br J Haematol 2001; 113: 951 ± 61 received revised accepted
April 30, 2008 June 25, 2008 June 29, 2008
Letter ¼ Planta Med 2008; 74: 1 ± 4
Bibliography DOI 10.1055/s-2008-1081339 Planta Med 2008; 74: 1±4 Georg Thieme Verlag KG Stuttgart ´ New York Published online n ISSN 0032-0943 Correspondence Karel SÏmejkal Department of Natural Drugs Faculty of Pharmacy University of Veterinary and Pharmaceutical Sciences Brno PalackØho 1±3 61242 Brno Czech Republic Phone: +420-5-4156-2839
[email protected]
