2 Department of Natural Medicinal Chemistry, China Pharmaceutical University, Nanjing 210009, China;
3 Department of Chemistry, Hong Kong University of Science and Technology, Hongkong, China
Melia toosendan Sieb. et Zucc. (Meliaceae), a Chinaberry tree, is distributed in southwest of China. Its fruits and bark are commonly used in traditional Chinese medicine for acesodyne and desinsection . Its chemical constituents reported in previous publications include highly functionalized limonoids such as meliacins, trichilinins, C-19/C-29- bridged acetals, ring C-seco limonoids, highly oxidized C-seco limonoids, spiro limonoids, and euphane- and tirucallane-type triterpenoids [2-5]. In our continued studies on the bioactive limonoids from M. toosendan, two new limonoids (compounds 1 and 2) together with toosendanin (compound 3) (Fig. 1), were isolated. X-ray crystallographic analysis established the absolute configuration of toosendanin (3). Herein, we report the isolation, structure elucidation, including the absolute configuration of toosendanin (3) and the biological activities of these compounds.
Compound 1, amorphous powder, was assigned the molecular formula C39H48O9 based on its HR-ESI-MS. The IR spectrum showed absorption peaks at 3414 cm–1 (hydroxyl group) and 1718 cm–1 (carbonyl group). The 1H and 13C NMR data of 1 (Table 1) demonstrated that compound 1 consisted of a cinnamoyl group [δH 6.49 (1H), 7.76 (1H), 7.52 (2H), 7.39 (2H), 7.39 (1H), δC 165.7 (CO), 119.6, 144.0, 134.5, 127.7, 129.0, 130.2], an acetyl group [(δH 1.88 (δC 21.3), δC 170.3], an ethoxy group [δH 1.15 (3H, t, J = 7.1 Hz), δC 15.0, δH 3.59 (1H, m), 3.17 (1H, m), δC 63.9], and a hydroxyl group. A furan ring [δH 6.42 (1H), 7.28 (1H), 7.30 (1H), δC 128.5, 110.6, 139.0, 142.7] was also apparent from the NMR spectra. By comparison the remaining NMR data with those limonoids isolated from M. toosendan before, a ring C-seco nimbolinin skeleton was assigned. The presence of the cinnamoyl group at the C-1 position was confirmed by the observation of the 3JCH connectivity in HMBC spectrum (Fig. 2) between H-1 (δH 4.95) and the carbonyl of the cinnamoyl group (δC 165.7). In addition, according to the HMBC spectrum, the acetyl group was placed at C-3, the hydroxyl group was attached to C-7, and the ethoxyl group was situated at C-12. The stereochemistry of 1 was established by a NOESY experiment (Fig. 2). The NOE correlations of CH3-29/H-3, CH3-29/H-6, and CH3-29/CH3-19 suggested the β-orientation of H-3 and thus the β-orientations of CH3-29, H-6, and CH3-19. The NOE correlation of CH3-19/H-1 revealed that H-1 was in the β-configuration. The NOE correlations between H-7/H-6 and H-7/CH3-30 suggested that H-7 was in the β-configuration, while the NOE correlations of H-15/H-16α and H-17/H-16α implicated α-configurations for H-15 and H-17. Observation of NOE effects between H-9/H-5, H-9/H-12, and H-9/H-15 indicated the α-configuration of H-12, which deduced a significant downfield shift for C-12 (δC 103.9). Thus, the structure of compound 1 was elucidated as 1α-cinnamoyl-3α-acetyl-7α- hydroxyl-12β-ethoxyl-nimbolinin, named 12-ethoxynimbolinin G.
Compound 2 was obtained as an amorphous powder, and its molecular formula was determined as C35H48O9 by HR- ESI-MS. The 1H NMR and 13C NMR spectra (Table 1) showed the presence of a tigloyl group [δH 7.01 (1H), 1.83 (3H), 1.91 (3H), δC 166.3 (CO), 137.0, 128.8, 14.5, 12.0], an acetyl group [(δH 2.02 (δC 21.0), δC 170.7), and an ethoxyl group [δH 1.23 (3H, t, J = 7.0 Hz), δC 15.0, δH 3.73 (1H, m), 3.33 (1H, m), δC 63.6]. On the basis of the NMR data, the same carbon skeleton as compound 1 was proposed. According to the HMBC spectrum, the acetyl group was placed at C-3, the tigloyl group was situated at C-7, and the hydroxyl group was linked to C-1. The NOESY experiment showed that the stereochemistry of compound 2 was consistent with compound 1. Therefore, the structure of compound 2 was established as 1α-hydroxyl- 3α-acetyl-7α-tigloyl-12β-ethoxyl-nimbolinin, named 12-ethoxynimbolinin H.
Compound 3 was obtained as amorphous powder. The molecular formula, C30H38O11, was determined by HR-ESI- MS (m/z 597.2418 [M + Na]+ (Calcd. for C30H38O11Na, m/z 597.2414)). It was determined as toosendanin based on the comparison of their spectral (Table 1) and physical data with those in the literature .
Toosendanin had been reported to be a mixture [6-7]. We also found that compound 3 was a mixture when it was in solution (two peaks in HPLC spectrum, probably representing two tautomers) and interestingly when in crystal status, it was proven to be a pure compound by X-ray.
To establish the absolute configuration of toosendanin, a single crystal X-ray diffraction study was carried out with Cu-Kα source and carefully measure of the anomalous X-ray scattering from the oxygen atom as well. The widely accepted FLACK parameter x (u) was –0.008(177)  whereas the newly accepted Hooft (Fleq) parameter y (s.u.) was determined as –0.003(0.068) from the refinement of over 1, 589 Bivioet pairs with Bijvoet-pair Analysis program in Platon, indicating that both FLACK and Hooft parameters proved the same handiness with low errors (esds) [9-10]. The absolute configuration at C-29 for toosendanin was determined to be R.
The stereo view of the molecule was shown in Fig. 3. The absolute configurations of chiral center for toosendanin were 1S, 3R, 4R, 5R, 7R, 8S, 9R, 10S, 12R, 13R, 14S, 15R, 17R, and 29R, as shown in Fig. 3. Both rings A and B were in distorted chair conformation while ring C was a twist boat conformation. Ring D was assumed as an envelope form with C(17) out of the mean plane passing through the other four atoms by 0.62Å. The five-membered ring D was trans fused to the six-membered ring C. The junctions between B/C and B/A were also trans.
Compounds 1−3 were evaluated for their cytotoxic activities against five human tumor cell lines: myeloid leukemia (HL-60), hepatocellular carcinoma (SMMC-7721), lung cancer (A-549), breast cancer (MCF-7), and colon cancer (SW480). However, only compound 1 showed cytotoxic activities against SMMC-7721 and MCF-7 cells, with IC50 values of 27.6 and 31.6 μmol·L−1, respectivelyExperimental General
Optical rotations were measured on a Perkin-Elmer 241 polarimeter(Waltham, Massachusetts, USA). IR spectra were recorded on a Perkin-Elmer 16 PC FT-IR spectrometer (Waltham, Massachusetts, USA). 1D and 2D NMR spectra were run on Bruker DRX-500 (Karlsruhe, Germany) and JEOL JNM-EX 400 spectrometers (Tokyo, Japan). HR-ESI-MS were obtained on a PE Biosystems Mariner System 5140 LC-MS spectrometer (Waltham, Massachusetts, USA). Silica gel (Merck, Darmstadt, Germany, Kieselgel 60, 230−400 mesh) was used for column chromatography. Preparative HPLC separation was performed on Waters Prep LC 4000 system (Milford, Massachusetts, USA) with a UV detector and an X-bridge C18 column (Waters, USA, 19 mm × 150 mm, 5 μm).Plant materials
The dried barks of Melia toosendan were collected from Wanxian, Sichuan Province, People's Republic of China, in October 2004 and authenticated by Prof. Minjian Qin (Department of Natural Medicinal Resources, China Pharmaceutical University, Nanjing, China). A voucher specimen (No. 24-88-53-3) was deposited in the Department of Natural Medicinal Chemistry, China Pharmaceutical University.Extraction and isolation
Dried barks of Melia toosendan (20 kg) were crushed and extracted thrice with 95% EtOH (40 × 1000 mL, 2 h each) under reflux. The extract was concentrated under vacuum to leave a black liquid (1167 g), which was suspended in water (3000 mL) and then extracted with petroleum ether (3000 mL, 1000 mL each time), chloroform (3000 mL, 1000 mL each time) and 1-butanol (3000 mL, 1000 mL each time) successively. The chloroform fraction (47 g) was subjected to silica gel column chromatography eluted with a gradient of chloroform−methanol (chloroform, 9 : 1, 8 : 2, 7 : 3, 6 : 4, 1 : 1, methanol, each 8000 mL) to yield fractions 1−10. Fraction 2 (3.0 g) was chromatographed on silica gel (4 × 42 cm, 230− 400 mesh) eluted with a gradient of petroleum ether-acetone (90 : 10, 88 : 12, 85 : 15 and 80 : 20, each 1500 mL) to afford fractions 2.1−2.9. Fraction 2.4 (40 mg) was purified by reversed phase preparative HPLC using a gradient of increasing acetonitrile (60%−90%) in water at 18 mL·min−1 for 20 min to give 1 (tR = 11.5 min, 15 mg) and 2 (tR = 13.2 min, 12 mg). Fraction 7 (4.2 g) was separated by silica gel column chromatography (4 cm × 42 cm, 230-400 mesh) eluted with CHCl3− acetone (65 : 35, 60 : 40, 55 : 45, each 1500 mL), followed by purification on Sephadex LH-20 [2 cm × 80 cm, CHCl3-MeOH (1 : 1), 200 mL], which was further purified by reversed-phase preparative HPLC using a gradient of increasing acetonitrile (28%−29%) in water at 18 mL·min−1 for 35min to yield 3 (tR = 17.9 min, 84 mg).
Compound 1 amorphous powder; [α]D25 −10.4 (c 0.12, CHCl3); IR (KBr) νmax = 3414, 2927, 1718, 1245, 1054, 770 cm–1; 1H- and 13C NMR (CDCl3), see Table 1; HR-ESI-MS m/z 683.3198 [M + Na]+ (Calcd. for C39H48O9Na, 683.3196).
Compound 2 amorphous powder; [α]D25 +9.6 (c 0.10, CHCl3); IR (KBr) νmax = 2924, 1719, 1459, 1254, 1055 cm–1; 1H- and 13C NMR (CDCl3), see Table 1; HRESIMS m/z 635.3187 [M + Na]+ (Calcd for C35H48O9Na, 635.3184).
Compound 3 amorphous powder; mp 251-253℃; [α]D25 –31.0 (c 0.75, MeOH); IR (KBr) νmax 3560, 1750, 1510, 880 cm–1; 1H and 13C NMR (CD3OD) data, see Table 1; HR-ESI-MS m/z 597.2418 [M + Na]+ (Calcd. for C30H38O11Na, 597.2414).X-ray analysis of toosendanin
Crystals suitable for X-ray diffraction were grown from a mixture solution of methanol and water and mounted on a glass fibre for diffraction experiment. Intensity data were collected on an Oxford Diffraction Xcalibur™S Ultra (Oxford, UK) with CCD Area Detector at 100K with Enhance Ultra Cu-Kα radiation (λ = 1.54178Å). Lattice determination, data collection, and reduction were carried out using CrysAlisPro 171.32.5 (Agilent, USA). Absorption corrections were performed using built-in SADABS program of the CrysAlisPro program suite. The structures were solved by the direct methods and refined by full-matrix least squares on F2 using the SHELXTL version 6.10 (Bruker, Germany). All non-hydrogen atoms were refined anisotropically. The hydrogen atoms were introduced at their geometric positions and refined as riding atoms.
Crystal data of 3: C30H38O11·3 (H2O) Mr = 629.66, crystal size 0.40 × 0.25 × 0.20 mm3; triclinic P1; T = 100(2) K; a = 7.3677(2) Å, b = 9.1845(3) Å, c =12.4965(4)Å, α= 103.453(3)°, β = 92.979(3)°, γ = 112.608(3)°. V = 749.74(4)Å3; Dc = 1.395 Mg/m3; Z = 1; F(000) = 337; μ = 0.931 mm–1. A total of 8082 reflections were collected in the range 3.68° < θ < 66.99°, with 4148 independent reflections [R(int) = 0.0160]; completeness (θ = 66.99°) = 96.1%; max. and min. transmission 1.00 and 0.84; data/restraints/parameters: 4148/4/422; goodness-of-fit on F2 = 1.035; R1 = 0.0425, wR2 (all data) = 0.1197; largest difference peak and hole = 0.239 and –0.282 e/Å3. (CCDC number 676895)Cytotoxicity assay
The five tumor cell lines (HL-60 human myeloid leukemia, SMMC-7721 hepatocellular carcinoma, A-549 lung cancer, MCF-7 breast cancer, and SW480 colon cancer) were bought from Cell Resource Center of Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences (Shanghai, China). Cells were cultured in DMEM supplemented with 10% heatinactivated fetal bovine serum (FBS), 100 U·mL−1 penicillin and 100 μg·mL−1 streptomycin at 37 ℃ in an atmosphere containing 5% CO2 and 95% air. The cytotoxic effect against all the cells was determined using the MTT assay . Briefly, 100 μL of adherent cells was seeded into each well of 96-well cell culture plates and allowed to adhere for 24 h before drug addition, while suspended cells were seeded just before drug addition to an initial density of 5 × 104 cells/mL. Each tumor cell line was exposed to the tested compounds dissolved in DMSO at various concentrations in triplicate for 72 h, with cisplatin as a positive control and cells without treatment were used as a control. Formazan crystals were solubilized by addition of DMSO, and optical density (OD) value was measured at 490 nm. The percentage of cell growth inhibition was calculated based on a comparison with untreated cells.
All the experiments were run in triplicate. The data were expressed as means ± SEM (standard errors). Statistical analysis was performed by unpaired Student's t-test for two comparisons. P < 0.05 was considered statistically significant.Acknowledgement
We thank Prof. LIANG Xiao-Tian for his interest and valuable suggestions.
Chen SK, Chen BY, Li H. Flora Republicae Popularis Sinicae[M]. Beijing: Science Press, 1997: 102-103.
Nakatani M. Limonoids from Melia toosendan (Meliaceae) and their antifeedant activity[J]. Heterocycles, 1999, 5(1): 595-609.
Tada K, Takido M, Kitanaka S. Limonoids from fruit of Melia toosendan and their cytotoxic activity[J]. Phytochemistry, 1999, 51(6): 787-791. DOI:10.1016/S0031-9422(99)00115-6
Zhang Q, Shi Y, Liu XT, et al. Minor limonoids from Melia toosendan and their antibacterial activity[J]. Planta Med, 2007, 73(12): 1298-1303. DOI:10.1055/s-2007-981618
Sang YS, Zhou CY, Lu AJ, et al. Protolimonoids from Melia toosendan[J]. J Nat Prod, 2009, 72(5): 917-920. DOI:10.1021/np800669c
Xie JX, Yuan AX. The structure of iso-chuanliansu isolated from Chinese medicine−the bark of Melia toosendan and Melia azedarach[J]. Acta Pharm Sin, 1985, 20(3): 188-192.
Shu GX, Liang XT. A correction of the structure of Chuancliansu[J]. Acta Chim Sin, 1980, 38(2): 196-198.
Flack HD. On enantiomorph-polarity estimation[J]. Acta Cryst, 1983, 39(6): 876-881. DOI:10.1107/S0108767383001762
Flack HD, Bernardinelli G. Reporting and evaluating absolute-structure and absolute-configuration determinations[J]. J Appl Crystallogr, 2000, 33(4): 1143-1148. DOI:10.1107/S0021889800007184
Hooft RWW, Straver LH, Spek AL. Determination of absolute structure using Bayesian statistics on Bijvoet differences[J]. J Appl Crystallogr, 2008, 41(1): 96-103. DOI:10.1107/S0021889807059870
Mosmann T. Rapid colorimetric assay for cellular growth and survival: Application to proliferation and cytotoxicity assays[J]. J Immunol Methods, 1983, 65(1): 55-63.