Novel indole and quinoline alkaloids from Melodinus yunnanensis

Introduction

Monoterpenoid indole alkaloids originate from the condensation of tryptophan with secologanin to produce strictosidine, which further alters by rearrangement to yield a dozen subgroups.¹ Some of the remarkable quinoline alkaloids, such as quinine and camptothecin which are well known for their antimalarial and anticancer properties, respectively, have been proposed to arise by rearrangement of monoterpenoid indole alkaloids. In a possible route for quinine biosynthesis, the cleavage of a N₁-C₂ bond in the indole heterocyclic ring could generate new amine and keto functions. A new quinoline heterocycle would then be formed by combining this N-1 amine with a C-5 aldehyde produced by a tryptamine sidechain cleavage, producing cinchonidinone.² Unlike quinine, the proposed biosynthesis of camptothecin includes a C₂–C₇ double bond oxidation to yield two carbonyls and an aldoltype condensation between C₂ and C₆ to form a quinoline ring. An in vivo tracer experiment has supported the prediction that the quinoline moiety originates from tryptophan.³ Also, melodinine B, a possible key intermediate of indole to quinoline alkaloids, has been reported.⁴ Plants of the genus Melodinus have been shown to be good sources of monoterpenoid indole and quinoline alkaloids.⁵ During our search for novel and bioactive monoterpenoid indole alkaloids from the family Apocynaceae, some representative skeletons and cytotoxic compounds were reported from the genera Alstonia⁶ and Melodinus.^4,7 This paper describes skeletons and cytotoxic compounds were reported from the genera Alstonia⁶ and Melodinus.^4,7 This paper describes the isolation, structural determination, proposed biosynthesis, and cytotoxic activities of 4 novel alkaloids (1–4) from M. yunnanensis.

Results and Discussion

Compound 1 was found to possess a molecular formula of C₁₉H₂₀N₂O, as evidenced by high resolution electron spray ionization mass spectra (HRESIMS) at m/z 293.1650, in combination with ¹H, ¹³C NMR, DEPT spectra, and appropriate for 11 degrees of unsaturation. The UV spectrum showed the presence of conjugated groups by showing maximum absorptions at 247 and 278 nm, and the IR spectrum indicated the presence of carbonyl and olefin groups (absorption bands at 1702 and 1610 cm^-1, respectively). In the ¹H NMR spectrum, two doublet (δ_H 7.55 and 7.13) and two triplet (δ_H 7.56 and 6.91) signals indicated that 1 was an unsubstituted indole alkaloid. In addition, signals for double bonds (δ_H 6.69 (d, J = 7.0 Hz, H-16), 5.42 (dd, J = 7.0, 1.4 Hz, H-17), 5.81 (m, H-14), and 5.64 (dd, J = 10.0, 1.4 Hz, H-15)), a methyl group (δ_H 0.62, t, H-18), and a methylene (δ_H 0.95 and 0.87, each 1H, m, H-19) were similar to those of ∆¹⁴-vincamenine (5).⁸ The ¹³C NMR and DEPT spectra of 1 showed signals of a methyl group (δ_C 8.5, q), four sp3 methylenes (δ_C 52.1, 53.1, 39.5, and 33.0), eight sp2 (δ_C 133.5, 126.7, 126.0, 125.0, 124.1, 120.3, 137.9, and 112.0) and one sp3 methines (δ_C 72.1), and three sp2 (δ_C 204.0, 159.8, and 120.8) and two sp3 quaternary carbons (δ_C 73.9 and 42.2). In comparison with those of 5, the four quaternary carbon signals at δ_C 131.0 (C-2), 134.8 (C-13), 129.2 (C-8), and 107.9 (C-7) of 5 were absent in 1, instead δ_C 204.0 (s), 73.9 (s), 159.8 (s), and 120.8 (s) were present in 1. In the heteronuclear multiple bond coherence (HMBC) spectrum of 1 (Figure 2), the correlation of δ_H 7.55 (1H, H-9) with δ_C 204.0 (s) suggested an indolone fragment, and the correlation of δ_H 6.69 (H-16), 1.88, 2.08 (H-6), and 3.10 (H-5) with δ_C 73.9 (s) led to the assignment of C-2 to a conjunct carbon of a spiral ring (Figure 1). Nuclear Overhauser Effect (NOE) correlation between H-21 and H-19 in the ROESY spectrum of 1 indicated three protons on the same side, a α-orientation identical to the configuration of its biosynthetic precursor, ∆¹⁴-vincamenine.⁹ Thus, the C-5 and 6 of the spiral ring (C) were on the upside of the planar indole configuration in a molecular model. Rings A, D, and E of 1 were same as those of 5, as indicated by the HMBC and ROESY spectra.

Open image in new window

Open image in new window

Compound 2 possessed a molecular formula C₂₀H₂₄N₂O₂ according to HRESIMS (m/z 325.1913, [M + H]⁺). The ¹H and ¹³C NMR spectra were very similar to those of 1, except for δ_C 85.5 (d), 36.2 (t), and an additional methoxyl δ_C 54.2 (q) in 2 instead of the olefinic signals at C-16/17 in 1. This suggested an added methoxyl group at C-16 in 2, which was further supported by the HMBC spectrum, with correlations of δ_H 5.04 (1H, t, J = 8.0 Hz, H-16) with C-2, 13, and 17. In the ROESY spectrum, NOE correlation of δ_H 5.04 (H-16) with δ_H 0.77 and 0.87 (2H, m, H-19), and of δ_H 3.39 (3H, s, OCH₃) with 2.19 (2H, m, H-6) placed the methoxyl as β. Other structural parts of 2 were identical to 1, as indicated by the HMBC and ROESY spectra.

Compound 3 possessed a molecular formula of C₂₀H₂₂N₂O₂, indicated by HRESIMS at m/z 323.1752 [M + H]⁺ in combination with ¹H, ¹³C NMR, and DEPT spectra. Its UV spectrum showed a conjugation pattern (absorptions at 230 and 292 nm) different from 1. In the ¹H NMR spectrum, four downfield protons (δ_H 8.34 (d, J = 8.4 Hz, H-9), 7.93 (d, J = 8.4 Hz, H-12), 7.68 (t, J = 8.4 Hz, H-11), and 7.37 (t, J = 8.4 Hz, H-10)) suggested a nonsubstituted quinolone rather than indole,¹⁰ further supported by the HMBC correlation of δ_H 8.34 (d, H-9) with a conjugated ketone signal at δ_C 176.0 (s, C-7). Detailed comparison of the NMR data of 2 and 3, indicated fused 6/5 rings (B and C) in 3 instead of corresponding spiral 5/5 rings in 2. Moreover, a methoxyl in β-orientation was supported by the NOE correlation of H-16 with H-19 in its ROESY spectrum.

The HRESIMS (m/z 327.1708, [M + H]⁺) of 4 established the molecular formula of C₁₉H₂₂N₂O₃. The NMR pattern was similar to melohenine B,⁴ except for olefinic signals [δ_C 122.7 (d) and 133.4 (d) and δ_H 5.83 (1H, m) and 5.74 (1H, d, J = 10.0 Hz)] observed in 4. The double bond was assumed to be at C-14/15 from comparison of spectral data from the two compounds and was further supported by the HMBC spectrum of 4, with correlations of δ_C 39.6 (s, C-20) with δ_H 5.83 (1H, m, H-14) and 5.74 (1H, d, H-15). NOE correlations of H-21 with H-19 and H-16 supported a 16β-OH for 4, identical to melohenine B.

The possible biosynthetic relationships of these new compounds in which they were derived from a common precursor 5 was proposed here (Figure 3). Different oxidation processes may have produced two kinds of intermediates, from which 4 was isolated. Further rearrangement then formed two new skeletons, including the spiro-indolone alkaloids (1 and 2), and a quinolone alkaloid (3). To our knowledge, this is the first report of the co-occurrence of monoterpenoid indoles (1, 2, and 5), a quinoline (3), and their key intermediates (4) in same plant, supporting the biosynthesis of quinoline from indole and concurring with previous in vivo tracer experiments in the literature.

Open image in new window

Compounds 1–5 were evaluated for their cytotoxicity against five human cancer cell lines, MCF-7 breast, SMMC-7721 hepatocellular carcinoma, HL-60 myeloid leukemia, SW480 colon cancer, and A-549 lung cancer. Compound 2 showed cytotoxicity against all cell lines, with IC₅₀ values of 14.24, 19.08, 15.48, 13.29, and 40.0 μm, respectively, while the cisplatin control showed IC₅₀ values of 21.90, 15.24, 1.05, 19.92, and 9.40 μm. The other compounds were inactive (IC₅₀ > 40 μm).

Experimental Section

General Experimental Procedures. Optical rotations were measured with a Jasco P-1020 spectropolarimeter. UV spectra were recorded on a Shimadzu double-beam 210A spectrophotometer. IR (KBr) spectra were obtained on Bruker Tensor 27 infrared spectrophotometer. ¹H, ¹³C and 2D NMR spectra were recorded on a Bruker avance Ⅲ-600 and AV-400 MHz NMR spectrometer with TMS as internal standard. MS data were obtained on API Qstar Pulsar Ⅰ spectrometer. C18

silica gel (20–45 μm) was bought from Fuji Chemical Ltd., Japan. MPLC was employed Büchi pumps system coupled with glass column (15 × 230 and 26 × 460 mm, respectively, C18 silica gel). HPLC was performed using Waters 600 pumps coupled with analytical and semipreparative Sunfire C18 columns (4.6 × 150 and 19 × 150 mm, respectively). The HPLC system employed a Waters 2996 photodiode array detector and a Waters fraction collector Ⅱ.

Plant Materials. Leaves and twigs of M. yunnanensis collected in Apr. 2009 in Honghe of Yunnan Province, P. R. China, and identified by Dr. De-Shan Deng. Voucher specimen (Cai091106) was deposited in the State Key Laboratory of Phytochemistry and Plant Resources in West China, Kunming Institute of Botany, Chinese Academy of Sciences.

Extraction and Isolation. Dried and powdered leaves and twigs of M. yunnanensis (40 kg) were extracted three times with methanol (MeOH) at room temperature and the solvent evaporated in vacuo. The residue was dissolved in 0.3% aqueous hydrochloric acid, and the solution subsequently basified using ammonia water to pH 9–10. The basic solution was partitioned with EtOAc, producing an aqueous and EtOAc phase. The resulting EtOAc fraction (205 g) was collected and then subjected to column chromatography over silica gel and eluted with a chloroform-acetone gradient (1/0 to 3/1, v/v) to afford five fractions (Ⅰ–Ⅶ), which were concentrated in vacuo. Fraction Ⅱ (5 g) was further chromatographed using a petroleum ether-acetone gradient (19/1 to 9/1, v/v) as the eluent to yield 8 subfractions, Ⅱ-1–Ⅱ-8. Compound 4 (46 mg) was crystallized from Ⅱ-8 (90 mg). Subfraction Ⅱ-1 (112 mg) was separated by semi-preparative reversed-phase C₁₈-HPLC on a Sunfire column (19 × 250 mm) with a gradient flow of 70–75% aqueous MeOH to yield 2 (5 mg). Ⅱ-3 (120 mg) was further purified on a same semi-preparative column with a gradient flow of 65–80% aqueous MeOH to afford 5 (107 mg). Similar semi-preparative column separations with gradient flows of 60–75% aqueous MeOH were used to fractionate Ⅱ-4 (90 mg) and Ⅱ-6 (100 mg) to produce 1 (4 mg) and 3 (5 mg), respectively.

Meloyunine A (1): white powder; [α]_D²⁰ + 223 (c 0.16, MeOH); UV (MeOH) λ_max (log ε) 247 (3.37), 278 (3.21) nm; IR (KBr) ν_max 3431, 1702, 1610 cm^-1; ¹H and ¹³C NMR data, Table 1; ESIMS m/z 293 [M + H]⁺; positive ion HRESIMS m/z 293.1650 (calcd for C₁₉H₂₁N₂O [M + H]⁺, 293.1653).

Meloyunine B (2): white powder; [α]_D²⁰ + 181 (c 0.15, MeOH); UV (MeOH) λ_max (log ε) 246 (3.37), 280 (2.80) nm; IR (KBr) ν_max 3433, 1703, 1613 cm^-1; ¹H and ¹³C NMR data, Table 1; ESIMS m/z 325 [M + H]⁺; positive ion HRESIMS m/z 325.1913 (calcd for C₂₀H₂₅N₂O₂ [M + H]⁺, 325.1916).

Meloyunine C (3): white powder; [α]_D²⁰ + 141 (c 0.13, MeOH); UV (MeOH) (log ε) λ_max 230 (3.46), 292 (3.15) nm; IR (KBr) ν_max 3431, 1708, 1623, 1592 cm^-1; ¹H and ¹³C NMR data, Table 1; ESIMS m/z 323 [M + H]⁺; positive ion HRESIMS m/z 323.1752 (calcd for C₂₀H₂₃N₂O₂ [M + H]⁺, 323.1759).

14, 15-Didehydromelohenine B (4): white powder; [α]_D²⁰ + 133 (c 0.06, MeOH); UV (MeOH) (log ε) λ_max 228 (3.35), 298 (3.14) nm; IR (KBr) ν_max 3483, 1686, 1668, 1598 cm⁻¹; ¹H and ¹³C NMR data, Table 1; ESIMS m/z 327 [M + H]⁺; positive ion HRESIMS m/z 327.1708 (calcd for C₁₉H₂₃N₂O₃ [M + H]⁺, 327. 1708).

Cytotoxicity Assay. Five human cancer cell lines, MCF-7 breast, SMMC-7721 hepatocellular carcinoma, HL-60 myeloid leukemia, SW480 colon cancer, and A-549 lung cancer, were used in the cytotoxic assay. Cells were cultured in RPMI-1640 or in DMEM medium (Hyclone, USA), supplemented with 10% fetal bovine serum (Hyclone, USA) in 5% CO₂ at 37 ℃. The cytotoxicity assay was performed according to the MTT (3-(4, 5-dimethylthiazol-2-yl)-2, 5-diphenyl tetrazolium bromide) method in 96-well microplates.¹¹ Briefly, 100 μL of adherent cells was seeded into each well of 96-well cell culture plates and allowed to adhere for 12 h before addition of test compounds, while suspended cells were seeded just before drug addition with initial density of 1 × 10⁵ cells/mL. Each tumor cell line was exposed to the test compound at concentrations of 0.0625, 0.32, 1.6, 8, and 40 μM in triplicates for 48 h, with cisplatin (Sigma, USA) as positive control. After compound treatment, cell viability was detected and a cell growth curve was graphed. IC₅₀ values were calculated by Reed and Muench's method.¹²

Notes

Electronic Supplementary Material

Supplementary material is available in the online version of this article at http://dx.doi.org/10.1007/s13659-011-0001-0 and is accessible for authorized users.

Acknowledgments

This work was supported in part by the National Natural Science Foundation of China (2107298), the 973 Program of Ministry of Science and Technology of P. R. China (2009CB522300), and Chinese Academy of Sciences (KSCX2-EW-R-15 and XiBuZhiGuang Project).