2172 CHEMISTRY & BIODIVERSITY – Vol. 4 (2007) Phenolic Derivatives with Free-Radical-Scavenging Activities from Ixeridium gracile (DC.) Shih by Xue-Mei Ma a ) b ), Yong Liu a ), and Yan-Ping Shi* a ) a ) Key Laboratory for Natural Medicine of Gansu Province, Lanzhou Institute of Chemical Physics, Chinese Academy of Sciences, Lanzhou 730000, P. R. China (phone: þ 86-931-4968208; fax: þ 86-931-8277088; e-mail: shiyp@lzb.ac.cn) b ) Graduate University of Chinese Academy of Sciences, Chinese Academy of Sciences, Beijing 100049, P. R. China Phytochemical investigation of the Tibetan medicinal plant Ixeridium gracile resulted in the isolation and identification of twelve flavonoids and two coumarins, compounds 1 – 14, the dimeric 92a,3a-epoxy5,7,3’,4’-tetrahydroxyflavan-(4b ! 8)-epicatechin; (1) being a new constituent. The free-radical-scavenging potentials of different extract fractions as well as of the pure compounds towards the DPPH (1,1diphenyl-2-picrylhydrazyl) radical were evaluated, and are discussed in terms of structure – activity relationship (SAR). The flavonoids were found to be the major constituents contributing to the freeradical-scavenging activity of I. gracile, but the high concentration of coumarins additionally contributed to the observed activity. Introduction. – Ixeridium gracile, a herbaceous perennial plant belonging to the Compositae family, is the only Ixeridium species distributed in the Sitsang region. This plant is used as a representative Tibetan herbal medicine, and it is also valued as a delicious and nutritional potherb [1]. Up to now, we have carried out systematic phytochemical studies on several medicinal herbs used in traditional Tibetan medicine, which grow only at altitudes of 2500 – 4000 m on the Qinghai-Tibet Plateau, and which were found to be abundant in flavonoids [2 – 4]. Numerous studies have indicated that flavonoids have antioxidant, anticarcinogenic, antiviral and anti-inflammatory activities [5], especially antioxidant activity being widely reported [6 – 8]. Accordingly, flavonoids have been considered to be important natural antioxidants for a long time, some constituents playing an important role in the prevention of lipid peroxidation and cardiovascular disease [9] [10]. In our previous studies, we focused on the chemical constituents of Tibetan herbs and obtained some plant-derived natural chemicals, especially flavonoids. However, neither the biological activities of these extracts nor the structures of the chemical constituents had been outlined in detail. In this paper, we report the results of a thorough phytochemical study on the extracts and constituents of I. gracile and their free-radical-scavenging potentials, as well as a qualitative structure – activity relationship (QSAR) study. From the EtOH extract of I. gracile, we isolated a total of 14 compounds, constituents 1 – 14, including a novel flavan dimer, 92a,3a-epoxy-5,7,3’,4’-tetrahydroxyH 2007 Verlag Helvetica Chimica Acta AG, ZJrich CHEMISTRY & BIODIVERSITY – Vol. 4 (2007) 2173 flavan-(4b ! 8)-epicatechin; (1) 1). Flavan dimers are a class of natural or synthetic products consisting of two flavan-3-ol units. These units are typically connected through either C(4)C(8) (as in 1) or through C(4)C(6), but a few other types of linkages have also been reported [11] [12]. Results and Discussion. – 1. Structure Elucidation. Compound 1 was obtained as an optically active, red oil ([a]20 D ¼ þ 14 (c ¼ 0.6, MeOH)). Its IR spectrum showed absorption bands for OH groups (3380 cm  1), aromatic rings (1604, 1570, 1513 cm  1), and a methylene moiety (1470 cm  1). HR-ESI-MS showed the quasi-molecular [M þ H] þ peak at m/z 577.1335, suggesting the molecular formula C30H24O12 , which was supported by the 1H- and 13C-NMR (DEPT) data (Tables 1 and 2, resp.). In the downfield region of the 1H-NMR spectrum of 1 (Table 1), eight singlets at d(H) 8.18 – 9.33 indicated the presence of eight phenolic OH groups. Nine aromatic signals were found, their spin-coupling patterns indicating two 1,3,4-trisubstituted aromatic rings (d(H) 6.99 (d, J ¼ 2.0), 6.82 (d, J ¼ 8.0), 6.73 (dd, J ¼ 8.0, 2.0), 7.05 (d, J ¼ 2.0), 6.74 (d, J ¼ 8.0), and 6.85 (dd, J ¼ 8.0, 2.0 Hz)) and one 1,2,3,5-tetrasubstituted aromatic ring (d(H) 5.91 (d, J ¼ 2.4), 5.85 (d, J ¼ 2.4 Hz)). The above 1) For systematic names, see Exper. Part. 2174 CHEMISTRY & BIODIVERSITY – Vol. 4 (2007) Table 1. 1H-NMR Data of 1. At 400 MHz in ( D6 )DMSO; d in ppm, J in Hz. Trivial atom numbering. Ring H-Atom d( H) Ring H-Atom d( H) A 5-OH HC(6) 7-OH HC(8) 8.18 (s) 5.91 (d, J ¼ 2.4) 9.11 (s) 5.85 (d, J ¼ 2.4) D 5-OH HC(6) 7-OH 9.33 (s) 5.98 (s) 9.29 (s) B HC(2’) 3’-OH 4’-OH HC(5’) HC(6’) 6.99 (d, J ¼ 2.0) 8.96 (s) 8.93 (s) 6.82 (d, J ¼ 8.0 ) 6.73 (dd, J ¼ 8.0, 2.0) E HC(2’) 3’-OH 4’-OH HC(5’) HC(6’) 7.05 (d, J ¼ 2.0) 8.84 (s) 8.71 (s) 6.74 (d, J ¼ 8.0) 6.85 (dd, J ¼ 8.0, 2.0) C HC(3) HC(4) 3.83 (d, J ¼ 4.4) 4.30 (d, J ¼ 4.4) F HC(2) HC(3) 3-OH CH2(4) 4.79 (d, J ¼ 1.0) 4.07 – 4.10 (m) 4.67 (d, J ¼ 4.4) 2.54 (dd, J ¼ 17.2, 2.0, Hax ), 2.79 (dd, J ¼ 17.2, 4.8, Heq ) Table 2. 13 C-NMR Data of 1. At 100 MHz in ( D6 )DMSO; d in ppm. Trivial atom numbering. Ring C-Atom d(C ) HMBC a ) Ring C-Atom d(C ) HMBC a ) A 5 6 7 8 9 10 156.15 96.37 156.55 94.50 152.85 102.62 4, 6 ( A ) D 154.89 94.46 150.42 105.56 152.85 101.07 4 (F) 6 (A) 4 (C), 8 ( A ) 3, 4 (C ), 6, 8 ( A) 5 6 7 8 9 10 3, 4 (C ) 4 (F) 4 (F) B 1’ 2’ 3’ 4’ 5’ 6’ 130.85 114.93 144.31 145.41 114.70 117.84 3, 2’, 5’, 6’ ( B ) 6’ ( B ) 2’, 5’ ( B) 5’, 6’ ( B) 6’ ( B ) 2’, 5’ ( B) E 1’ 2’ 3’ 4’ 5’ 6’ 130.10 115.02 144.66 144.53 114.72 118.82 2, 3 ( F ) 2 (F) 2’, 5’ ( E ) 5’, 6’ ( E ) 6’ (E ) 2 (F) C 2 3 4 98.50 66.15 27.72 3, 4 (C ), 2’, 6’ (B ) 4 (C) 3 (C) F 2 3 4 79.28 64.49 29.34 a 3, 4 ( F ), 2’, 6’ ( E ) 2, 4 ( F ) 2, 3 ( F ) ) HMBC Correlations from 13C to 1H. 1 H-NMR spectroscopic data suggested that 1 was a flavan dimer. The two trisubstituted aromatic rings were attributed to the B- and E-rings of the flavan structure, and the 1,2,3,5-tetrasubstituted aromatic ring was assigned to the A-ring. An additional singlet at d(H) 5.98 (1 H) was assigned to HC(6) 2 ) of the D-ring, which was found to be pentasubstituted, based on the cross-peak between d(H) 5.98 and d(C) 94.46 in the 1 H,13C-HMQC experiment and comparison with literature data [13]. 2) Arbitrary atom numbering. CHEMISTRY & BIODIVERSITY – Vol. 4 (2007) 2175 In the 1H-NMR spectrum of 1, the signals of four aliphatic H-atoms (d(H) 2.54 (dd, J ¼ 17.2, 2.0), 2.79 (dd, J ¼ 17.2, 4.8), 4.07 – 4.10 (m), 4.79 (d, J ¼ 1.0 Hz)) were typical of a flavan-3-ol moiety with 2,3-cis configuration (so-called epicatechin type) [14 – 16]. A pair of doublets (d(H) 3.83, 4.30 (J ¼ 4.4 Hz each)) as an AB-type spin system in the heterocyclic region were assigned to HC(3) and HC(4), respectively, of the C-ring, based on HMQC cross-peaks between d(H) 3.83 and d(C) 66.15, and between d(H) 4.30 and d(C) 27.72. The coupling constant (J ¼ 4.4 Hz) was in accordance with a relative 3,4-trans-configuration [16]. All the 1H-NMR assignments were confirmed by 1 H,1H-COSY correlations. The 13C-NMR (DEPT) spectrum of 1 (Table 2) indicated 30 C-atoms: one CH2 group, 13 CH groups (including nine aromatic and three oxygenated ones), as well as 16 quaternary C-atoms (15 being aromatic). The resonances at d(C) 79.28, 64.49, and 24.34 supported the presence of the proposed epicatechin structural moiety, and those at d(C) 98.50 and 66.15 pointed to an epoxide. So, compound 1 was composed of a 2,3epoxy-5,7,3’,4’-tetrahydroxyflavan and an epicatechin moiety, which was further confirmed by the following key HMBC correlations (Fig. 1): C(2) of ring C with HC(2) and HC(6) of ring B; C(10) of ring A with HC(3) and HC(4) of ring C; C(2) of ring F with HC(2) and HC(6) of ring E; and C(10) of ring D with HC(3) and HC(4) of ring F. The two substructures could be linked by long-rang correlations between both HC(3) and HC(4) (ring C) and C(8) (ring D), and between HC(4) (ring C) and the atoms C(7), C(8), and C(9) (ring D), respectively. Fig. 1. Selected HMBC (H ! C) correlations for 1 A comparison of the spectroscopic data of 1 with those of 9flavan-3-ol (4a ! 8)pelargonidin 3-O-b-glucopyranoside;, including a 4a-substituent at ring C [17], showed that the chemical shifts for HC(4) at the C-ring were different. In compound 1, the signal was shielded by ca. 0.8 ppm. In the case of the known compound aesculitannin G, a flavan dimer with a C(4)C(8) linkage, the orientation of HC(4) is b, on the basis of the coupling constant (J  0 Hz) of HC(4). In compound 1, the corresponding J(3,4) value was 4.4 Hz, thus indicating an a-orientated HC(4) at the C-ring [18]. Similarly, comparison of the 1H-NMR data of 1 with those of A-type procyanidin showed that the signals for HC(4) (C ring) were basically identical, which corroborated that 1 had a 4b-substituent at the C-ring [16]. From these data, the structure of 1 was elucidated as 92a,3a-epoxy-5,7,3’,4’-tetrahydroxyflavan-(4b ! 8)epicatechin;. 2176 CHEMISTRY & BIODIVERSITY – Vol. 4 (2007) The known compounds 2 – 14 isolated from I. gracile were identified by 1H-,13C(DEPT), and 2D-NMR (1H,1H-COSY, gHMQC, gHMBC) experiments, and by direct comparison of their spectroscopic data with those previously reported. They were identified as 5-hydroxyflavanone (2) [19], 5,7-dihydroxyflavanone (3) [20], 7hydroxyflavanone (4) [21], 7-methoxyflavanone (5) [22], 2’,4’-dihydroxychalcone (6) [23], 2’,4’-dihydroxydihydrochalcone (7) [23], (3R)-7,2’-dihydroxy-3’,4’-dimethoxyisoflavan (8) [24], 7-hydroxycoumarin (9) [25], 5,8-dihydroxy-7-methoxycoumarin (10) [26], quercetin (11) [27], kaempferol (12) [28], quercetin 3-O-galactoside (13) [27], and luteolin 7-O-glucoside (14) [29]. 2. Free-Radical-Scavenging Activity. The quenching of the DPPH (1,1-diphenyl-2picrylhydrazyl) radical is a very convenient method for screening small antioxidant molecules, because the reaction can be observed visually using common TLC and dotblot techniques, and can be analyzed by simple spectrophotometric or chromametric assays [2] [30]. The free-radical-scavenging activity of selected extract fractions and of compounds 2 – 14 were, thus, determined by DPPH radical-scavenging assay, and the results are shown in Table 3. As can be seen, almost all of these samples exhibited considerable scavenging activity. Qualitative analysis revealed that the petroleum ether (PE)-, CHCl3-, AcOEt-, and BuOH-soluble fractions, as well as the alcoholic extract all displayed free-radical-scavenging activity. Quantitative analysis of the different fractions revealed that the activity increased from weak to strong in the order PE < BuOH < AcOEt < CHCl3 . Table 3. Free-Radical-Scavenging Activities of Different Extract Fractions and Isolated Compounds from Ixeridium gracile. For details regarding the DPPH assay, see Exper. Part. All values are means from triplicate experiments. Fraction/compound IC50  S.D. [mg/ml] Fraction/compound IC50  S.D. [mg/ml] Petroleum ether Chloroform Ethyl acetate Butanol Alcoholic extract 2 3 4 5 6 7 8 11.14  0.01 0.56  0.01 1.31  0.02 5.57  0.01 7.18  0.01 157.1  0.10 140.6  0.23 168.2  0.15 179.3  0.11 132.3  0.16 251.6  0.10 277.1  0.10 9 10 11 12 13 14 Q3OG a ) K3OG a ) K3OA a ) Ascorbic acid a ) BHTa ) 107.9  0.41 95.3  0.80 12.8  0.20 52.8  0.21 41.2  0.41 37.1  0.28 39.7  0.48 110  0.28 125  0.55 10.8  0.30 20.2  0.32 a ) Positive controls: Q3OG, quercetin 3-O-glucopyranoside; K3OG and K3OA, kaempferol 3-Ogalactopyranoside and -arabinopyranoside, resp.; BHT, 9butylated hydroxytoluene;. The major chemical constituents of the PE-soluble fraction were flavanones, which contributed to the free-radical-scavenging activity of this fraction; the same was true for the many flavanones and coumarins in the CHCl3-soluble fraction and the abundant flavones in the AcOEt-soluble fraction. The flavone glycosides as the main chemical CHEMISTRY & BIODIVERSITY – Vol. 4 (2007) 2177 constituents in the BuOH-soluble extract were less active than the constituents in the AcOEt-solube one. The free-radical-scavenging activities of the isolated compounds 2 – 8 (Table 3) were lower than those of 11 – 14, which was attributed to the lower number of OH groups as well as hydrogenation of the C¼C bond between C(2) and C(3) of 2 – 8. Further analysis indicated that 11 – 14 possessed higher scavenging activities than 12, an effect associated with their ortho-dihydroxy B-ring structures, conferring higher radical stability and participating in electron delocalization, as reported previously for other systems [31]. The higher free-radical-scavenging activity of 11 (IC50 ¼ 12.8 mm) than 13 (41.2 mm) confirmed that O-glycosylation in 3-position has a negative effect on activity. In summary, the free-radical-scavenging activity of natural flavonoids seems to be mainly influenced by four factors (Fig. 2). First, the number and location of aromatic OH groups, especially of the ortho-dihydroxy B ring [32], is important for activity. Second, increasingly hydrophobic substituents decrease the activity [33]. Third, a C¼O group at C(4) is important. Fourth, OH groups at C(3) and C(5), but especially at C(3), are essential for activity [34]. Fig. 2. Structure – activity relationship (SAR) for the free-radical-scavenging activity of flavanes Up to now, very few studies have been reported on the antioxidant activities of coumarins. Paya et al. [35] carried out some studies on the superoxide-scavenging capacity on a series of synthetic or plant-derived coumarins with different substitution patterns. They found that only 7,8-dihydroxylated coumarins are active. Piao et al. [36] found that the free-radical-scavenging activities of furanocoumarins towards DPPH are correlated with the number of phenolic OH groups [36]. Our analysis of compounds 9 and 10 also indicates that coumarins are less active than flavonoids. However, their different concentrations also plays an important role in terms of the overall activity, especially in case of the CHCl3-soluble fraction. Nevertheless, the above analysis shows that the main free-radical-scavenging constituents of I. gracile are flavonoids. The high abundance of these compounds in Tibetan herbs might be attributed to the unique plateau environment, and must play an important role in the plant;s ecological adaptation to the harsh conditions at high altitude. To make Tibetan medicines available to more people, several researchers have qualitatively and quantitatively analyzed many plants used in local folk medicine to determine their specific names, pharmacological components, functions and effects, and their proper use and dosage. As a result, Tibetan medicine has become more standardized and scientific. From medicinal and health points of view, the unique Tibetan herb I. gracile seems worth of further research. This work was supported by the National Natural Science Foundation of China (No. 20372029 and 20475057) and the National Natural Science Foundation of Gansu Province (No. 3ZS051-A25-077). 2178 CHEMISTRY & BIODIVERSITY – Vol. 4 (2007) Experimental Part General. DPPH was obtained from Sigma-Aldrich Chemical Co. (St. Louis, MO, USA). Petroleum ether (PE) had a b.p. range of 60 – 908. TLC: silica gel GF254 (10 – 40 mm; Qingdao Marine Chemical Factory). Column chromatography (CC): silica gel (200 – 300 mesh; Qingdao). Optical rotations: PerkinElmer 341 polarimeter. IR Spectra: Nicolet NEXUS-670 FT-IR spectrometer; in cm  1. NMR Spectra: Varian INOVA-400 FT-NMR spectrometer; d in ppm rel. to Me4Si, J in Hz. HR-ESI-MS: Bruker APEXII mass spectrometer; in m/z. Plant Material. Ixeridium gracile (DC.) Shih was bought from the Tibetan Hospital of Qinghai Province, P. R. China, in August 2004, and was identified by Dr. Huan-Yang Qi. A voucher specimen (No. ZY03002) was deposited at the Key Laboratory for Natural Medicine of Gansu Province, Lanzhou Institute of Chemical Physics, Chinese Academy of Sciences, P. R. China. Preparation of Extract Fractions. Air-dried and ground whole plants of I. gracile (3.0 kg) were extracted repeatedly with 85% aq. EtOH (7  3 d) at r.t. The combined extracts were evaporated to dryness under reduced pressure. The residue (120 g) was suspended in H2O (1000 ml), and extracted successively with PE (8  1500 ml), CHCl3 (2  1500 ml), AcOEt (6  1500 ml), and BuOH (4  1500 ml). Compound Isolation. a) The PE-soluble extract (60 g) was subjected to CC (600 g SiO2 ; PE/acetone 30 : 1 ! 1 : 1) to afford four fractions (Fr. 1 – 4). Fr. 1 (30 : 1 ! 20 : 1) was further purified by CC (SiO2 ; PE/ CHCl3 2: 1) to yield 2 (10 mg) and 3 (9 mg). Fr. 2 (15 : 1 ! 10 : 1) was repeatedly purified by CC (SiO2 ; PE/Et2O 4 : 1) to give 4 (20 mg). Fr. 3 (8 : 1 ! 5 : 1) was purified by CC (SiO2 ; PE/acetone 6 : 1) to afford 5 (9 mg). Fr. 4 (5 : 1 ! 1 : 1) was subjected to CC (SiO2 ; PE/acetone 5 : 1) to give three subfractions (Fr. 4a – 4c). Fr. 4a was repeatedly purified by CC (SiO2 ; CHCl3/acetone 15 : 1) to provide 6 (10 mg). Fr. 4b was also subjected to CC (SiO2 ; PE/AcOEt 3 : 1) to yield 7 (17 mg). Fr. 4c was purified by CC (SiO2 , PE/ AcOEt 4 : 1) to give 8 (10 mg). b) The CHCl3-soluble extract (6 g) was subjected to CC (60 g SiO2 ; PE/ AcOEt 20 : 1 ! 1 : 1) to give four fractions (Fr. A – D). Fr. A (20 : 1 ! 12 : 1) was repeatedly chromatographed (SiO2 ; CHCl3/acetone 16 : 1) to provide 6 (5 mg) and 7 (12 mg). Fr. B (10 : 1 ! 8 : 1) was subjected to CC (SiO2 ; CHCl3/AcOEt 10 : 1) to give 8 (5 mg). Fr. C (8 : 1 ! 5 : 1) was purified by CC (SiO2 ; CHCl3/acetone 15 : 1) to give crude 9 (23 mg), which was further purified by CC (SiO2 ; CHCl3/ AcOEt 12 : 1) to provide pure 9 (20 mg). Fr. D (5 : 1 ! 1 : 1) was separated by CC (SiO2 ; CHCl3/acetone 15 : 1) to provide 10 (30 mg). c) The AcOEt-soluble extract (30 g) was subjected to CC (300 g SiO2 ; CHCl3/MeOH 20 : 1 ! 1 : 1) to give four subfractions (Fr. I – IV). Fr. I (20 : 1 ! 12 : 1) was chromatographed (polyamide gel; MeOH/H2O 4 : 1) to give 11 (40 mg). Fr. II (10 : 1 ! 8 : 1) was purified by CC (SiO2 ; CHCl3/MeOH 5 : 1) to provide 12 (6 mg). Fr. III (7 : 1 ! 5 : 1) was subjected to CC (polyamide gel; MeOH/H2O 2 : 1) to give 13 (20 mg). Fr. IV (5 : 1 ! 1 : 1) was also subjected to CC (polyamide gel; MeOH/H2O 4 : 1) to yield 1 (20 mg). d) The BuOH-soluble extract (15 g) was subjected to CC (150 g SiO2 ; CHCl3/MeOH 20 : 1 ! 1 : 1) to afford four fractions. The fraction eluted at a gradient of 5 : 1 ! 2 : 1 was purified by CC (polyamide gel; MeOH/H2O 4 : 1) to yield 14 (50 mg). 92a,3a-Epoxy-5,7,3’,4’-tetrahydroxyflavan-(4b ! 8)-epicatechin; ( ¼ (2R*,3R*)-2-(3,4-Dihydroxyphenyl)-8-[(1aS*,7R*,7aR*)-1a-(3,4-dihydroxyphenyl)-1a,7a-dihydro-4,6-dihydroxy-7H-oxireno[b] [1]benzopyran-7-yl]-3,4-dihydro-2H-1-benzopyran-3,5,7-triol; 1). Red oil. [a]20 D ¼ þ 14 (c ¼ 0.6, MeOH). IR (film): 3380, 1604, 1570, 1513, 1470. 1H- and 13C-NMR: see Tables 1 and 2, resp. HR-ESI-MS: 577.1335 ([M þ H] þ , C30H25Oþ12 ; calc. 577.1346). 5-Hydroxyflavanone (2). Colorless needles (CHCl3 ). 1H-NMR (400 MHz, CDCl3 ): 12.12 (s, 5-OH); 7.33 – 7.44 (m, HC(2’) to HC(6’)); 5.46 (dd, J ¼ 13.2, 2.8, HC(2)); 3.02 (dd, J ¼ 16.8, 13.2, Ha C(3)); 2.86 (dd, J ¼ 16.8, 2.8, Hb C(3)). 13C-NMR (100 MHz, CDCl3 ): 191.97 (C(4)); 166.32 (C(5)); 163.52 (C(9)); 137.64 (C(1’)); 134.65 (C(7)); 128.88 (C(5’)); 128.72 (C(4’)); 128.52 (C(3’)); 127.46 (C(6’)); 127.12 (C(2’)); 123.16 (C(8)); 120.18 (C(6)); 103.41 (C(10)); 79.77 (C(2)); 44.08 (C(3)). 5,7-Dihydroxyflavanone (3). Colorless crystals (PE/CHCl3 ). 1H-NMR (400 MHz, CDCl3 ): 12.14 (s, 5-OH); 7.32 – 7.42 (m, HC(2’) to HC(6’)); 6.26 (d, J ¼ 2.4, HC(8)); 6.18 (d, J ¼ 2.4, HC(6)); 5.42 (dd, J ¼ 13.2, 2.8, HC(2)); 3.02 (dd, J ¼ 16.8, 13.2, Ha C(3)); 2.86 (dd, J ¼ 16.8, 2.8, Hb C(3)). 13 C-NMR (100 MHz, CDCl3 ): 191.81 (C(4)); 166.47 (C(5)); 165.45 (C(9); 163.46 (C(7)); 137.74 (C(1’)); 128.98 (C(5’)); 128.82 (C(4’)); 128.52 (C(3’)); 126.46 (C(6’)); 126.12 (C(2’)); 103.69 (C(10)); 96.80 (C(6)); 95.58 (C(8)); 79.70 (C(2)); 43.32 (C(3)). CHEMISTRY & BIODIVERSITY – Vol. 4 (2007) 2179 7-Hydroxyflavanone (4). Colorless crystals (CHCl3 ). 1H-NMR (400 MHz, (D6 )acetone): 7.56 (d, J ¼ 8.8, HC(5)); 7.35 – 7.44 (m, HC(2’) to HC(6’)); 6.56 (dd, J ¼ 8.8, 2.4, HC(6)); 6.43 (d, J ¼ 2.4, HC(8)); 5.55 (dd, J ¼ 13.2, 2.8, HC(2)); 3.02 (dd, J ¼ 16.8, 3.2, Ha C(3)); 2.84 (dd, J ¼ 16.8, 2.8, Hb C(3)). 13C-NMR (100 MHz, (D6 )acetone): 189.52 (C(4)); 164.58 (C(7)); 163.37 (C(9)); 139.45 (C(1’)); 128.49 (C(6’)); 128.46 (C(5’)); 128.42 (C(4’)); 128.26 (C(3’)); 127.56 (C(2’)); 126.27 (C(5)); 113.95 (C(10)); 110.37 (C(6)); 102.57 (C(8)); 79.52 (C(2)); 43.70 (C(3)). 7-Methoxyflavanone (5). Colorless crystals (PE/AcOEt). 1H-NMR (400 MHz, CDCl3 ): 7.84 (d, J ¼ 8.8, 2.4, HC(6)); 7.34 – 7.44 (m, HC(2’) to HC(6’)); 6.61 (d, J ¼ 2.4, HC(8)); 6.48 (d, J ¼ 8.8, HC(5)); 5.44 (dd, J ¼ 13.2, 2.8, HC(2)); 3.80 (s, 5-MeO); 3.02 (dd, J ¼ 16.8, 13.2, Ha C(3)); 2.84 (dd, J ¼ 16.8, 2.8, Hb C(3)). 13C-NMR (100 MHz, CDCl3 ): 166.16 (C(7)); 163.50 (C(9)); 138.76 (C(1’)); 128.83 (C(6’)); 128.75 (C(5’)); 128.74 (C(4’)); 128.51 (C(3’)); 128.50 (C(5)); 126.13 (C(2’)); 119.58 (C(4)); 113.88 (C(10)); 110.26 (C(6)); 100.88 (C(8)); 79.99 (C(2)); 44.31 (C(3)). 2’,4’-Dihydroxychalcone (6). Yellow needles (PE/acetone). 1H-NMR (400 MHz, (D6 )acetone): 7.88 (d, J ¼ 15.6, a-H); 7.85 (d, J ¼ 15.6, b-H); 7.33 – 7.43 (m, HC(2) to HC(6)); 6.44 (d, J ¼ 8.8, HC(6’)); 6.34 (dd, J ¼ 8.8, 2.4, HC(5’)); 6.17 (d, J ¼ 2.4, HC(3’)). 13C-NMR (100 MHz, (D6 )acetone): 166.50 (C(4’)); 163.23 (C(2’)); 144.43 (b-C); 134.77 (C(1)); 131.94 (C(6’)); 130.45 (C(4)); 128.99 (C(6)); 128.86 (C(2)); 128.52 (C(3)); 126.14 (C(5)); 120.31 (a-C); 114.29 (C(1’)); 107.88 (C(5’)); 103.71 (C(3’)). 2’,4’-Dihydroxydihydrochalcone (7) . Colorless needles (PE/acetone). 1H-NMR (400 MHz, (D6 )acetone): 12.70 (s, 2’-OH); 9.99 (s, 4’-OH); 7.72 (d, J ¼ 8.8, HC(6’)); 7.12 – 7.24 (m, HC(2) to HC(6)); 6.35 (dd, J ¼ 8.8, 2.8, HC(5’)); 6.28 (d, J ¼ 2.8, HC(3’)); 3.23 (m, a-H); 2.04 (m, b-H). 13 C-NMR (100 MHz, (D6 )-acetone): 177.50 (C¼O)); 164.56 (C(2’)); 164.33 (C(4’)); 140.79 (C(1)); 132.07 (C(6’)); 132.07 (C(6)); 128.12 (C(5)); 127.98 (C(4)); 127.95 (C(3)); 125.56 (C(2)); 112.33 (C(1’)); 107.60 (C(5’)); 102.24 (C(3’)); 38.78 (a-C); 29.13 (b-C). (3R)-7,2’-Dihydroxy-3’,4’-dimethoxyisoflavan (8). Colorless needles (PE/acetone). 1H-NMR (400 MHz, (D6 )acetone): 8.56 (s, 7-OH); 8.17 (s, 2’-OH); 6.81 (d, J ¼ 8.4, HC(6’)); 6.46 (d, J ¼ 8.4, HC(5’)); 6.33 (dd, J ¼ 8.4, 2.4, HC(6)); 6.26 (d, J ¼ 2.4, HC(8)); 6.25 (d, J ¼ 8.4, HC(5)); 4.22 (m, Ha C(2)); 3.93 (m, Hb C(2)); 3.79 (s, 4’-MeO); 3.76 (s, 3’-MeO); 3.41 (m, HC(3)); 2.80 (m, Ha C(4)); 2.08 (m, Hb C(4)). 13C-NMR (100 MHz, (D6 )acetone): 156.51 (C(7)); 154.97 (C(9)); 151.67 (C(4’)); 147.96 (C(2’)); 135.86 (C(3’)); 129.97 (C(5)); 121.40 (C(6’)); 120.55 (C(1’)); 112.99 (C(10)); 107.76 (C(6)); 103.16 (C(5’)); 102.55 (C(8)); 69.30 (C(2)); 59.78 (3’-MeO); 55.03 (4’-MeO); 31.94 (C(3)); 29.86 (C(4)). The structure was further secured by means of 1H,1H-COSY and gHMBC experiments. 7-Hydroxycoumarin (9). Colorless needles (acetone). 1H-NMR (400 MHz, (D6 )acetone): 7.86 (d, J ¼ 8.8, HC(4)); 7.47 (d, J ¼ 8.8, HC(5)); 6.81 (dd, J ¼ 8.8, 2.4, HC(6)); 6.72 (d, J ¼ 2.4, HC(8)); 6.15 (d, J ¼ 8.8, HC(3)). 13C-NMR (100 MHz, (D6 )acetone): 161.08 (C(2)); 161.02 (C(9)); 155.67 (C(7)); 143.77 (C(4)); 130.32 (C(5)); 114.66 (C(6)); 112.63 (C(10)); 111.37 (C(3)); 101.92 (C(8)). 5,8-Dihydroxy-7-methoxycoumarin (10). Colorless needles. 1H-NMR (400 MHz, (D6 )acetone): 9.58 (s, 8-OH); 9.45 (s, 5-OH); 7.86 (d, J ¼ 9.2, HC(4)); 6.78 (s, HC(6)); 6.19 (d, J ¼ 9.2, HC(3)); 3.73 (s, 7-MeO). 13C-NMR (100 MHz, (D6 )acetone): 160.55 (C(2)); 152.04 (C(5)); 145.3 (C(9)); 145.07 (C(4)); 139.32 (C(8)); 139.25 (C(7)); 111.84 (C(3)); 110.24 (C(10)); 100.02 (C(6)). Quercetin (11). Yellow crystals (MeOH). 1H-NMR (400 MHz, (D6 )DMSO): 12.48 (s, 5-OH); 10.80 (s, 7-OH); 9.60 (s, 3-OH); 9.36 (s, 4’-OH); 9.32 (s, 3’-OH); 7.51 (dd, J ¼ 8.4, 2.4, HC(6’)); 7.41 (d, J ¼ 2.4, HC(2’)); 6.90 (d, J ¼ 8.4, HC(5’)); 6.66 (d, J ¼ 2.4, HC(6)); 6.44 (d, J ¼ 2.4, HC(8)). 13C-NMR (100 MHz, (D6 )DMSO): 175.86 (C(4)); 163.90 (C(7)); 160.73 (C(9)); 156.15 (C(5)); 146.81 (C(4’)); 145.07 (C(2)); 145.07 (C(3’)); 135.75 (C(3)); 121.97 (C(1’)); 120.00 (C(6’)); 115.62 (C(5’)); 103.03 (C(10)); 98.20 (C(6)); 93.37 (C(8)); 115.07 (C(2’)). Kaempferol (12). Yellow crystals (MeOH). 1H-NMR (400 MHz, (D6 )DMSO): 12.38 (s, 5-OH); 8.11 (dd, J ¼ 8.8, 2.4, HC(5’)); 8.08 (dd, J ¼ 8.8, 2.4, HC(3’)); 7.00 (dd, J ¼ 8.8, 2.4, HC(6’)); 6.99 (dd, J ¼ 8.8, 2.4, HC(2’)); 6.50 (d, J ¼ 2.0, HC(6)); 6.34 (d, J ¼ 2.0, HC(8)). 13C-NMR (100 MHz, (D6 )DMSO): 175.21 (C(4)); 163.78 (C(7)); 160.53 (C(9)); 158.92 (C(4’)); 156.33 (C(5)); 145.75 (C(2)); 135.17 (C(3)); 129.01 (C(2’)); 121.81 (C(6’)); 121.66 (C(1’)); 115.41 (C(3’)); 114.86 (C(5’)); 102.56 (C(10)); 97.68 (C(6)); 93.04 (C(8)). Quercetin 3-O-Galactoside (13). Yellow crystals (MeOH). 1H-NMR (400 MHz, (D6 )DMSO): 12.38 (s, 5-OH); 7.65 (dd, J ¼ 8.4, 2.4, HC(6’)); 7.51 (d, J ¼ 2.4, HC(2’)); 6.81 (d, J ¼ 8.4, HC(5’)); 6.46 (d, 2180 CHEMISTRY & BIODIVERSITY – Vol. 4 (2007) J ¼ 2.4, HC(6)); 6.36 (d, J ¼ 2.4, HC(8)); 5.37 (d, J ¼ 7.6, HC(1) of Gal); 3.8 – 4.8 (m, 6 H of Gal). 13 C-NMR (100 MHz, (D6 )DMSO): 177.50 (C(4)); 164.13 (C(7)); 161.24 (C(5)); 156.31 (C(9)); 156.24 (C(2)); 148.49 (C(4’)); 144.85 (C(3’)); 133.47 (C(3)); 122.03 (C(6’)); 121.09 (C(1’)); 115.93 (C(5’)); 115.19 (C(2’)); 103.93 (C(10)); 101.75 (C(1) of Gal); 98.68 (C(6)); 93.52 (C(8)); 75.10 (C(5) of Gal); 73.86 (C(3) of Gal); 71.18 (C(2) of Gal); 67.93 (C(4) of Gal); 60.14 (C(6) of Gal). Luteolin 7-O-Glucoside (14). Yellow crystals (MeOH/H2O). 1H-NMR (400 MHz, (D6 )DMSO): 6.43 (s, HC(2)); 6.73 (d, J ¼ 2, HC(6)); 6.79 (d, J ¼ 2, HC(8)); 7.40 (d, J ¼ 2.4, HC(2’)); 6.88 (d, J ¼ 8.4, HC(5’)); 7.44 (dd, J ¼ 8.4, 2.4, HC(6’)); 12.95 (s, 5-OH); 5.06 (d, J ¼ 8.8 Hz, HC(1) of Glc); 3.25 – 3.70 (m, 6 H of Glc). 13C-NMR (100 MHz, (D6 )DMSO): 182.02 (C(4)); 164.58 (C(7)); 163.06 (C(2)); 161.48 (C(5)); 157.12 (C(9)); 149.94 (C(4’)); 145.79 (C(3’)); 121.53 (C(1’)); 119.37 (C(6’)); 116.10 (C(5’)); 113.58 (C(2’)); 105.48 (C(3)); 103.33 (C(10)); 100.00 (C(1) of Glc); 99.65 (C(6)); 94.98 (C(8)); 77.23 (C(4) of Glc); 76.35 (C(3) of Glc); 73.15 (C(2) of Glc); 69.60 (C(5) of Glc); 60.67 (C(6) of Glc). Qualitative Radical-Scavenging Assay. The DPPH radical-scavenging assay was based on the method of Soler-Rivas et al. [30]. Stock solns. of the PE-, CHCl3-, AcOEt-, and BuOH-soluble fraction and the alcoholic extract (see Table 3) were prepared in MeOH at a concentration of 5 mg/ml each. The pos. controls were 9butylated hydroxytoluene; (BHT) and ascorbic acid in MeOH (1 mg/ml). The radicalscavenging capacities of the samples were tested by thin layer chromatography (TLC). A drop (2 ml) of each sample was put individually on the baseline of a TLC plate and left to dry for several minutes. Then, the TLC plate was immersed upside down in a 200 mm DPPH soln. in MeOH for several seconds. Samples appeared as yellow-colored spots against a purple background. The intensity of the yellow color depended upon the amount and nature of radical scavengers present in the samples. Quantitative Radical-Scavenging Assay. Spectrophotometric analysis was used to determine IC50 values of the pure compounds. The same controls were used as above, and stock solns. of 1 mg/ml in MeOH were prepared, and then serially diluted with MeOH to final concentrations of 10 – 500 mg/ml. An aliquot (1 ml) of test soln. was thoroughly mixed with an equal volume of 200 mm DPPH soln. This mixture was kept in the dark for 30 min. Then, the UV/VIS absorbance (A) at 517 nm was measured, and the scavenging activity was determined as percent scavenging activity ¼ ([Acontrol  Asample ]/Acontrol )  100. Here, IC50 is the amount of antioxidant necessary to decrease the initial concentration of the DPPH radical by 50%, as determined through extrapolation from linear-regression analysis. All experiments were performed in triplicate (n ¼ 3), mean values being reported. REFERENCES [1] Tibet Plateau Scientific Expedition Team of the Chinese Academy of Sciences, 9Flora Xizangica;, Science Press, Beijing, 1985, Vol. 4, p. 961. [2] Q. X. Wu, Y. Li, Y. P. Shi, J. Asian Nat. Prod. Res. 2006, 8, 391. [3] J. X. Liu, X. N. Wei, R. H. Lu, Y. P. Shi, Chin. Tradit. Herbal Drugs 2006, 37, 338. [4] A. M. Yang, X. Liu, R. H. Lu, Y. P. Shi, Chin. Tradit. Herbal Drugs 2006, 37, 25. [5] B. H. Havsteen, Pharmacol. Therapeut. 2002, 96, 67. [6] R. Torres, F. Faini, B. Modak, F. Urbina, C. Labbe, Phytochemistry 2006, 67, 984. [7] Y. L. Huang, P. Y. Yeh, C. C. Shen, C. C. Chen, Phytochemistry 2003, 64, 1277. [8] S. SJzgeÅ, A. H. MeriÅli, P. J. Houghton, B. SubukÅu, Fitoterapia 2005, 76, 269. [9] M. G. L. Hertog, P. C. H. Hollmann, Eur. J. Clin. Nutr. 1996, 50, 63. [10] S. Das, J. P. N. Rosazza, J. Nat. Prod. 2006, 69, 499. [11] D. Y. Xie, R. A. Dixon, Phytochemistry 2005, 66, 2127. [12] J. Coetzee, L. Mciteka, E. Malan, D. Ferreira, Phytochemistry 2000, 53, 795. [13] T. Hatano, A. Yamashita, T. Hashimoto, Phytochemistry 1997, 46, 893. [14] F. Ferrari, F. D. Monache, R. A. D. Lima, Phytochemistry 1998, 47, 1165. [15] S. Morimoto, G. Nonaka, I. Nishioka, Chem. Pharm. Bull. 1986, 34, 633. [16] J. F. W. Burger, H. Kolodziej, R. W. Hemingway, J. P. Steynberg, D. A. Ypung, A. D. Ferreira, Tetrahedron 1990, 46, 5733. [17] T. Fossen, S. Rayyan, O. M. Andersen, Phytochemistry 2004, 65, 1421. CHEMISTRY & BIODIVERSITY – Vol. 4 (2007) [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] 2181 S. Morimoto, G. Nonaka, I. Nishioka, Chem. Pharm. Bull. 1987, 35, 4725. Z. B. Hu, S. X. Liao, S. L. Mao, H. P. Zhu, S. Xu, Acta. Pharm. Sin. 1999, 34, 132. J. K. Macleod, H. B. Rasmussen, Phytochemistry 1999, 50, 105. M. S. J. Simmond, W. M. Blamey, F. Delle Monache, G. B. M. Bettolo, J. Chem. Ecol. 1990, 16, 365. K. Martin, G. Rudolf, J. Chromatogr. 1990, 514, 147. D. Q. Yu, J. S. Yang, 9 Handbook of Analytical Chemistry;, Chemical Industry Press, Beijing, 1999, Vol. 7, p. 830. Z. Q. He, J. Findjay, J. Nat. Prod. 1991, 54, 810. N. J. Cussans, T. N. Huckerby, Tetrahedron 1975, 31, 2719. D. Q. Yu, J. S. Yang, 9 Handbook of Analytical Chemistry;, Chemical Industry Press, Beijing, 1999, Vol. 7, p. 845. X. H. Zhu, H. M. Shi, Z. D. Ming, Tradit. Pat. Med. 2004, 26, 137. B. D. Sloley, L. J. Urichuk, P. Morley, J. Durkin, J. J. Shan, P. K. Pang, R. T. Coutts, J. Pharm. Pharmacol. 2000, 52, 451. J. C. Lu, X. Z. Feng, Q. S. Sun, H.-W. Lu, M. Masanobu, K. Sugahara, D. Ma, Y. Sagara, H. Kodama, Clin. Chim. Acta 2002, 316, 95. C. Soler-Rivas, J. C. Espin, H. J. Wichers, Phytochem. Anal. 2000, 11, 1. P. Pier-Giorgio, J. Nat. Prod. 2000, 63, 1035. W. Bors, W. Heller, C. Michel, M. Saran, 9Antioxidants in Therapy and Preventive Medicine;, Plenum Press, New York, 1990, pp. 165 – 170. F. Conforti, G. A. Statti, R. Tundis, F. Menichini, P. Houghton, Fitoterapia 2002, 73, 479. W. Bors, W. Heller, C. Michel, M. Saran, Methods Enzymol. 1990, 186, 343. M. Paya, P. A. Goodwin, B. De Las Heras, J. R. Hoult, Biochem. Parmacol. 1994, 48, 445. X. L. Piao, I. H. Park, S. H. Baek, H. Y. Kim, M. K. Park, J. H. Park, J. Ethnopharmacol. 2004, 93, 241. Received December 22, 2006