(±)-Evodiakine, A Pair of Rearranged Rutaecarpine-Type Alkaloids From Evodia rutaecarpa

Abstract (±)-Evodiakine (1a and 1b), a pair of rearranged rutaecarpine-type alkaloids with an unprecedented 6/5/5/7/6 ring system, were isolated from the nearly ripe fruits of Evodia rutaecarpa. Separation of the enantiomers have been achieved by chiral HPLC column. The structures of (±)-evodiakine were unambiguously elucidated by 1D and 2D NMR spectra, mass spectrometry, and single-crystal X-ray diffraction. Their absolute configurations were determined by comparison of experimental and calculated electronic circular dichroism spectra. A hypothetical biogenetic pathway for (±)-evodiakine was also proposed. Compounds 1a, 1b, and the racemate (1) were tested for their cytotoxic and anti-inflammatory activities. Graphical Abstract Electronic supplementary material The online version of this article (doi:10.1007/s13659-016-0113-7) contains supplementary material, which is available to authorized users.


Results and Discussion
Evodiakine (1) (Table 1) implied the possibility of a dihydroindoline derivative for 1 [7]. In addition, the proton signals at d H 6.56 (d, J = 7.7 Hz), 7.08 (t, J = 7.7 Hz), 6.81 (t, J = 7.7 Hz), 7.33 (d, J = 7.7 Hz); and d H 7.58 (d, J = 8.0 Hz), 7.74 (t, J = 8.0 Hz), 7.43 (t, J = 8.0 Hz), 7.72 (d, J = 8.0 Hz) displayed characteristic of two ortho-disubstituted benzenoid rings. The 13 C NMR, DEPT, and HSQC spectra suggested that 1 exhibited seven other carbon signals due   . These spectroscopic data suggested that 1 was an alkaloid with polyheterocyclic systems. The structure of 1 was established by the further 2D NMR data analysis. In the 1 H-1 H COSY spectrum (Fig. 3), the cross peaks of H-9/H-10/H-11/H-12 and H-16/H-17/H-18/H-19 confirmed the presence of the two H-bearing structural fragments (a and c) and consistent with the 1 H NMR spectrum. Additionally, the 1 H-1 H COSY correlations of H 2 -5/H 2 -6 suggested the existence of a CH 2 CH 2 alkyl fragment (b). The fragment of a together with the HMBC correlations from OH-7 (d H 6.49) to C-2, C-7, and C-8, from N 1 -H (d H 6.00) to C-2, C-7, C-8, and C-13, from H-9 with C-8 and C-11, and from H-12 to C-2, C-10, and C-13 revealed the presence of a 2,3-disubstituted indolin-3ol moiety (i) in 1. The HMBC correlations of H 2 -5 with C-2, C-6, and C-7 and of H 2 -6 with C-2, C-5, C-7, and C-8 demonstrated that the indolin-3-ol moiety (i) was fused with a pyrrolidine ring (ii) via C-2 and C-7. In addition, the rest structural fragment (iii) was established by the HMBC correlations of H-16 with C-15, C-17, C-20, and C-21, of H-19 with C-15, C-17, and C-21, and of H 3 -22 with C-3 and C-15. Furthermore, the HMBC correlation of N 1 -H with C-3, as well as the downfield chemical shift of C-2 at d C 91.2, suggested that the carbonyl group (C-3) was connected to the C-2 of the indolin-3-ol moiety (i). Finally, the absence of an N-4 proton resonance in the 1 H NMR spectrum combined with the molecular composition suggested that C-21 was linked to N-4 of the pyrrolidine moiety (ii). Thus, the structure of 1 was identified as a pentacyclic alkaloid by fusing indole and the secopyrroloquinazolone rings [9] (Fig. 3). This deduction was further confirmed by an X-ray diffraction experiment using Mo Ka radiation (Fig. 4).
Evodiakine (1) was optically inactive, [a] D 24.8 & 0 (c 0.12, MeOH), indicating that it was obtained as a racemate. Subsequent HPLC separation of 1 on a chiral column prior to HPLC separation yielded two compounds, 1a and 1b (Fig. 5). However, the isolated compounds showed opposite optical rotation, and their ECD spectra displayed mirror curves as shown in Fig. 6. This confirmed the successful separation of enantiomers, (1)-evodiakine (1a) and (-)-evodiakine (1b). To secure unambiguous confirmation of the absolute configuration of compounds 1a and 1b, the  ECD calculation at the B3LYP/6-31G** level in Gaussian 03 program package was carried out, which provided vindication of their configuration [10]. In the 200-400 nm region, the calculated ECD spectra of compounds 1a and 1b were consistent with the experimental ECD spectra of (1)-1 and (-)-1, respectively. Thus, the absolute configuration of compound 1a was determined to be 2S,7S-evodiakine, as well as that of 1b was revealed as 2R,7Revodiakine. By comparison with known rutaecarpine-type alkaloids, compound (±)-1 was regarded as a rearranged rutaecarpine-type alkaloid with an unprecedented 6/5/5/7/6 ring system. From a biogenetic point of view, (±)-evodialine (1) would plausibly be derived from a common rutaecarpine-type alkaloid, evodianinine (2), via the sequence shown in Scheme 1. Compound 2 underwent isomerization and oxidation to produce intermediate B [11]. Then, the C-3/N-4 bond cleavage and the formation of a heptatomic ring accomplished via the key intermediate C [12]. The rings C and D rearrangement of B via attack of the NH in C onto the imine carbon yielded the structure D, which was followed by reduction of the C-5/C-6 double bond to produce compounds 1a and 1b [13,14].
Compounds 1a, 1b, and the racemate (1) were evaluated for their anti-inflammatory and cytotoxic activities, but none of them showed ability to inhibit NO production of LPS-stimulated RAW 264.7 macrophages, as well as cytotoxicity against HL-60, SMMC-7721, A-549, MCF-7, and SW-480 cancer cell lines with IC 50 values of more than 40 lM.

General
Melting points were obtained on an X-4 micro melting point apparatus. Optical rotations were measured with a Perkin-Elmer model 241 polarimeter. UV spectra were recorded using a Shimadzu UV-2401A spectrophotometer. IR spectra were determined on a Tensor-27 infrared spectrophotometer with KBr pellets. ECD spectra were obtained on a JASCO J-810 spectrophotometer. 1D and 2D NMR spectra were performed on Avance III-600 spectrometers with TMS as an internal standard. ESIMS and HREIMS were measured using an API-QSTAR-Pulsar-1 or VG Auto Spec-3000 instruments. X-ray diffraction was conducted using Bruker APEX DUO diffractometer with graphite-monochromated MoKa radiation. MPLC was performed on a Lisui EZ Purify III System including pump manager P03, detector modules P02, and fraction collector P01 (Shanghai Lisui Chemical Engineering Co., Ltd., Shanghai, China). Column chromatography (CC) was performed over silica gel (200-300 mesh, Qingdao Marine Chemical Co. Ltd., Qingdao, China), MCI gel (CHP 20P, 75-150 lm, Mitsubishi Chemical Corporation, Japan). Thin-layer chromatography (TLC) was carried out on silica gel GF 254 on glass plates (Qingdao Marine Chemical Co. Ltd.) using various solvent systems and spots were visualized by spraying improved Dragendorff's reagent to the silica gel plates.

Plant Materials
The dried and nearly ripe fruits of E. rutaecarpa were purchased from the Kunming Ju-Hua village pharmaceutical sale market, Yunnan province, P. R. China.

Extraction and Isolation
The dried and near ripe fruits of E. rutaecarpa (20.0 kg) were extracted with MeOH (3 9 10 L) at room temperature for 24 h each time. The MeOH extracts were evaporated under reduced pressure to give a residue, which was dissolved in 1% HCl, acidified to pH 2, and then partitioned with EtOAc (3 9 4 L). The acidic solution was Scheme 1 Plausible biogenetic pathway of (?)-1a and (-)-1b basified using saturated Na 2 CO 3 to pH 10, followed by exhaustive extraction with CHCl 3 (3 9 4 L) to afford a extract (36 g

Inhibition of Nitric Oxide Production Assay
The inhibitory effects of the test compounds on NO production were evaluated based on a detection model with LPS-activated murine macrophage RAW264.7 cells, which was performed as described previously [15]. The concentration of NO in the cultured medium was measured indirectly by analysis of nitrite levels using the Griess reaction. MG-132 was used as a positive control.

Cytotoxicity Assay
The isolates were tested in vitro for their cytotoxicities to inhibit proliferation of five human tumour cell lines, HL-60, SMMC-7721, A-549, MCF-7, and SW480. Cell viability was assessed by conducting colorimetric measurements of the amount of insoluble formazan formed in mitochondrion of living cells according to the MTT method [16]. In brief, each cancer cell line was exposed to the compounds dissolved in DMSO at five different concentrations in triplicate for 48 h, with cis-platin as a positive control.