Monoterpenoid Indole Alkaloids from Inadequately Dried Leaves of Alstonia scholaris.

Six new indole alkaloids, named alstoniascholarines L-Q (1-6), together with nineteen known analogues were isolated from the inadequately dried leaves of Alstonia scholaris. Their structures were elucidated on the basis of extensive analysis of spectroscopic data and by comparison of their physical and spectroscopic data with the literature values. In addition, the new alkaloids were tested for their cytotoxic and neurite outgrowth-promoting activities.

Introduction

Plants of the genus Alstonia (Apocynaceae), which are usually shrubs or trees, grow mainly in the tropical regions of Africa and Asia. A. scholaris has been historically used to treat chronic respiratory diseases in ‘dai’ ethnopharmacy in Yunnan province, People’s Republic of China. Previous phytochemical and pharmacological studies on this species afforded a number of structurally diverse indole alkaloids [1-5] with various bioactivities, such as antineoplastic [6], antibacterial [7], anti-inflammatory and analgesic effects [8], and broncho-vasodilatory [9] activities. The leaves of A. scholaris are usually collected and dried by exposed to sunshine in an open yard. However, some leaves of A. scholaris might not be dried in time because of rainy days in Pu’er city. Then, the green color of these leaves might fade in couples of day even they were died finally. Whether these leaves can be used for medical raw material is still unknown, which encouraged us to carried out HPLC analysis of total alkaloids from the inadequately dried leaves of A. scholaris. The results indicated that picrinine, the major bioactive compound was decreased remarkably and more peaks were appeared in the HPLC fingerprint profile of inadequately dried leaves (see Supplementary data). Correspondingly, the anti-tussive efficacy reduced significantly (Table 1). In our continuing efforts to search for structurally interesting and bioactive indoles of this plant [10-15], the inadequately dried leaves of A. scholaris were investigated. As a result, six new indole alkaloids, alstoniascholarines L–Q (1–6), along with nineteen known analogues were isolated. We report herein the isolation, structural elucidation and bioactivities of alkaloids.

Table 1

Effect of the different alkaloid extracts on ammonia-induced cough in mice

Group	Dose (mg/kg)	Frequency of cough	Inhibition (%)
Control		31.3 ± 6.5	–
Codeine phosphate	30	7.0 ± 2.5**	77.6
Alkaloids from the dried leaves	20	13.1 ± 4.8**	58.0
	10	17.0 ± 5.6**	45.6
Alkaloids from the inadequately dried leaves	20	19.4 ± 3.9**	38.0
	10	22.8 ± 7.1*	27.2

Values expressed as mean ± SEM (n = 10), * P < 0.05 and ** P < 0.01 for comparison of treated groups with control

Results and Discussion

Alkaloid 1 was obtained as a white amorphous powder. The HREIMS spectrum showed a quasi-molecular ion peak at m/z 356.1366 [M]+ (calcd for C19H20N2O5, 356.1372) from which, in conjunction with the 13C NMR data, the molecular formula was determined to be C19H20N2O5, requiring 11 indices of hydrogen deficiency. The UV absorption maxima 204, 240, and 298 nm suggested the presence of an indole chromophore, while the IR spectrum of 1 indicated the presence of indolic amino (3440 cm−1), ester carbonyl (1748 cm−1), and aromatic ring (1612 and 1485 cm−1) functionalities, respectively. The 13C NMR and DEPT spectra (Table 2) for 1 revealed 19 carbon signals, including characteristic signals due to an indole ring [δC 109.2 (s, C-2), 51.2 (s, C-7), 136.7 (s, C-8), 124.3 (d, C-9), 121.1 (d, C-10), 130.2 (d, C-11), 110.3 (d, C-12), and 148.7 (s, C-13)], two ester carbonyls (δC 178.1 and 178.8), one oxygen-bearing quaternary carbon (δC 92.8), four methines (δC 68.9, 51.2, 49.1, and 32.3), three methylenes (δC 45.7, 44.2, and 20.7), and one methyl group (δC 17.5).

Table 2

1H and 13C NMR spectroscopic data for alkaloids 1–3 in CD3OD

Position	1 a		2 a		3 b
	δ C	δ H (J, Hz)	δ C	δ H (J, Hz)	δ C	δ H (J, Hz)
2	109.2		109.0		84.5
3	51.2	3.63 br d (5.7)	51.4	3.64 br d (5.8)	68.5	3.66 br d (3.1)
5	178.1		178.1		150.7	8.81 d (4.8)
6a	44.2	3.73 d (18.0)	44.0	3.78 d (18.0)	118.4	7.49 d (4.8)
6b		2.77 d (18.0)		2.75 d (18.0)
7	51.2		51.2		ndc
8	136.7		136.6		128.1
9	124.3	7.20 d (7.5)	124.2	7.20 d (7.5)	129.2	9.19 d (8.6)
10	121.1	6.79 t (7.6)	121.0	6.80 t (7.5)	128.0	7.69 t (8.4)
11	130.2	7.08 t (7.5)	130.2	7.08 t (7.7)	131.0	7.79 t (8.2)
12	110.3	6.62 d (7.8)	110.3	6.62 d (7.8)	129.7	8.06 d (8.3)
13	148.7		148.7		150.1
14a	20.7	1.92 ddd (8.0, 6.0, 2.0)	21.5	2.00 ddd (8.1, 6.1, 2.0)	26.0	2.22 d (13.8)
14b		1.77 dd (14.6, 4.3)		1.77 dd (14.5, 4.3)		1.98 dt (13.8)
15	32.3	2.83 m	34.3	2.62 m	42.5	3.11 m
16	49.1	3.23 d (11.5)	48.6	3.28 d (11.5)	55.7	3.36 d (7.9)
17	178.8		178.7		176.5
18	17.5	1.20 d (6.5)	17.4	1.20 d (6.5)	17.5	1.26 d (6.5)
19	68.9	3.72 q (6.5)	71.3	3.78 q (6.5)	70.4	3.81 q (6.5)
20	92.8		92.1		92.0
21a	45.7	3.06 d (13.3)	44.6	3.24 d (13.4)	54.2	2.94 d (12.8)
21b		2.68 d (13.3)		2.73 d (13.4)		2.79 d (12.8)
N-Me					46.2	2.65 s

aRecorded at 400 and 100 MHz

bRecorded at 600 and 150 MHz

cNot detected

The 1H–1H COSY spectrum of 1 disclosed the presence of three structural fragments, a (C-9–C-12), b (C-3–C-14–C-15–C-16), and c (C-18–C-19), as shown in Fig. 2. The position of functional groups and the skeleton of alkaloid 1 were assigned by its HMBC data. In the H MBC spectrum, the correlations of δH 3.73 and 2.77 (both d, J = 18.0 Hz, H2-6) with δC 109.2 (s, C-2), 178.1 (s, C-5) and 51.2 (s, C-7), indicated the presence of a five-membered lactone ring E which was formed by the connection of C-5 and C-2 via an oxygen atom. Besides, the HMBC correlations of H-16 with C-7 and of δH 3.63 (1H, br d, J = 5.7 Hz, H-3) with C-2 suggested the formation of a six-membered ring C. The relative downfield shifts of δC 51.2 (d, C-3) and 45.7 (t, C-21) required that they both be connected to a nitrogen atom. Likewise, the HMBC correlations of δH 2.68 (1H, d, J = 13.3 Hz, H-21b) with δC 32.3 (d, C-15), 92.8 (s, C-20) and C-3, and of δH 2.83 (1H, m, H-15) with C-20 established the occurrence of another six-membered ring. Moreover, considering the last one degree of unsaturation in 1, another ring should be constructed. The key HMBC correlations of H-16 with C-17 and C-20 revealed that another five-membered lactone ring was present between C-17 and C-20. Finally, a linkage of C-18/19/20 was deduced from HMBC correlations of Me-18 (δH 1.20, d, J = 6.5 Hz) with δC 68.9 (d, C-19) and C-20. On the basis of the aforementioned information, the planar structure of 1 was elucidated as an monoterpenoid indole alkaloid with a rare 6/5/6/6/5/5 fused ring system. The stereochemistry was then considered. In the ROESY spectrum, the correlations of H-3/H2-14, H2-14/H-15, and H-15/H-16 were observed, indicating that they should be placed on the same side. However, it still could not determine the configuration for such complicated ring system. According to the similarities between the NMR data of 1 and the recently reported bio-relationship (Scheme 1) of alstolactines A–C [14], the configuration of 1 was assigned as 2R, 3S, 7R, 15R, 16R, 19R, 20S. This was further confirmed by the comparison of its CD spectrum with that of alstolactine A (Fig. 3). It is worthy to note that alstolactines A–C were also isolated from the leaves of A. scholaris and their absolute configurations were determined by the X-ray diffraction. On the basis of the above results, the structure of alkaloid 1 was established to be as shown in Fig. 1 and named alstoniascholarine L.

Fig. 2

Key 1H-1H COSY () and HMBC () correlations for 1–4

Scheme 1

Putative biosynthetic pathway of alkaloids 1–3 from picrinine

Fig. 3

Experimental CD spectra of alstolactine A (red), 1 (blue) and 2 (green). (Color figure online)

Fig. 1

Structures of alkaloids 1–6

Alkaloid 2 was shown to have the same molecular formula of C19H20N2O5 as that of 1 based on its HREIMS spectrum (m/z 356.1358 [M]+, calcd for C19H20N2O5, 356.1372). The 1H and 13C NMR spectral data (Table 2) of 2 were almost identical with those of 1, except for the carbon signals of C-15, C-19, and C-21, which indicated that 2 could be a 19-epimer of 1. Following a similar analysis of the 1H, 13C and 2D NMR spectroscopic data as described before, the gross structure of alkaloid 2 was the same as that of 1. The well matched CD curves of alkaloids 1 and 2 (Fig. 3) proposed the absolute configurations of 1 and 2. The absolute configurations of 2R, 3S, 7R, 15R, 16R, 20S were those commonly accepted from biogenetic point of view. However, the configuration of C-19 was assigned as 19R rather than 19S based on its similar NMR data with that of alstolactine B [14]. Therefore, the structure of 2 was characterized as shown in Fig. 1 and named alstoniascholarine M.

Alkaloid 3 was determined to have the molecular formula of C20H22N2O4 from an HREIMS ion peak at m/z 354.1577 ([M]+). The 1H and 13C NMR spectroscopic data (Table 2) of 3 showed its structure resembling that of alstolactine A. The main difference between them was the presence of two olefinic carbon signals (δC 118.4 and 150.7) in alkaloid 3 and the absence of the lactone (δC 177.2) and methylene (δC 43.8) groups in 2. The HMBC correlation of H-5 (δH 7.49, d, J = 4.8 Hz) with C-7 (δC 84.5) and a 1H-1H COSY correlation between H-5 (δH 7.49) and H-6 (δH 8.81) allowed the location of the double bond was at C-5 and C-6. Other parts of 3 were identical to those of alstolactine A, as supported by its HSQC, HMBC, and COSY experimental data. The configuration of C-19 was determined to be R based on the chemical shifts of C-19 (δC 70.4) and C-20 (δC 92.0), which were in agreement to those of alstolactine A (δC 69.2 and 92.1) rather than those of alstolactine B (δC 71.7 and 91.4). The other configurations of 3 were assigned to be the same as those of 1 based on their same biogenetic pathway (Scheme 1) and similar NMR data. Accordingly, the structure of 3 was elucidated as shown in Fig. 1 and named alstoniascholarine N.

Biosynthetically, the three related alkaloids might be derived from the precursor picrinine, which was also isolated as a main chemical constituent in this experiment. A plausible biogenetic pathway for 1–3 suggested that a Hoffmann degradation led to the formation of a dihydrofuran intermediate A. Further oxidation would yield a lactone derivative B. Then, epoxidation at C-20–C-19 double bond, followed by ring opening could produce another lactonic F-ring. Subsequently, alkaloids 1 and 2 would finally be obtained via demethylation of the methyl N(4)-Me of alstolactines A and B. Coincidently, alkaloid 3 might also derive from intermediate A without oxidation and demethylation.

Alkaloid 4, isolated as a white amorphous powder, had the molecular formula C20H24N2O4, as established by its HRESIMS data (m/z 357.1810, [M + H]+, calcd for C20H25N2O4 357.1810), which indicated 10 degrees of unsaturation. Its IR spectrum displayed characteristic absorptions attributing to amino/hydroxyl (3417 cm−1), double bond (1678 cm−1), and aromatic ring (1620, 1440 cm−1) functionalities. The 1H NMR spectrum (Table 3) showed the presence of a 1, 2, 3-trisubstituted aromatic moiety due to the signals of three contiguous aromatic hydrogens (δH 6.89, d, J = 7.5 Hz, H-9; 7.14, t, J = 7.8 Hz, H-10; 6.82, d, J = 8.1 Hz, H-11) and two methyls (δH 3.79 and 1.32). The 13C NMR spectrum (Table 3) displayed a total of 20 carbon signals, which were classified as two methyls, four methylenes, seven methines, and seven quaternary carbons, respectively. The presence of conjugated ester functionality was supported by the observed carbon signals at δC 168.4 and 51.7, while the signals due to the two olefinic quaternary carbons at δC 165.6 (C-2) and 105.3 (C-16) are consistent with the presence of a β-anilinoacrylate moiety [8]. One downfield signal at δC 69.0 was associated with the presence of oxymethine. The COSY and HMQC data (Figs. 2, 3) disclosed the following partial structures, viz., NCH2CH2, NCHCH2CHCHCH2, and CHCHCH3, corresponding to the C-5–C-6, C-3–C-14–C-15–C-20-C–21, and C-20–C-19–C-18 fragments, respectively. The above NMR data and the partial structures from the COSY spectrum indicated a similarity N(4)-demethylalstogustine (25) except for the observed difference in the aromatic region. The replacement of H by OH at C-12 was verified by the HMBC correlations of δH 6.82 (H-10) and 6.74 (H-11) with δC 143.3 (C-12). The relative configurations at the various stereogenic centers were established from ROEs and by comparison of the 13C NMR spectra with the reported values. The ROESY spectrum (Fig. 4) displayed cross peaks of H-3/H-9, H-3/H-5b, H-3/H-6b, H-3/H2-14, H2-14/H-15 indicating the relative configurations of C-7 and C-3, which also in turn allowed the orientation of H-15 to be assigned as α. The preferred boat conformation adopted by the piperidine ring D could be deduced from the observed H-14a/H-21a ROE correlation [16]. Likewise, the key ROE correlations of H-9/H-6b and H-6a/H-20 suggested a β orientation for H-20. Characteristic chemical shifts of C-16 (δ 105.6) and C-19 (69.0) suggested that the configuration of C-19 in 4 should be R,which was consistent with that of alstogustine (δC 105.3, 68.8) [17], rather than that of 19-epi-alstogustine (δC 104.8, 70.6) [17]. On the basis of the above evidences, the structure of 4 was deduced to be as shown in Fig. 1 and named as alstoniascholarine O.

Table 3

1H and 13C NMR spectroscopic data for alkaloids 4–6 in CD3ODa

Position	4		5		6
	δ C	δ H (J inHz)	δ C	δ H (J in Hz)	δ C	δ H (J inHz)
2	165.6		164.8		166.2
3	61.8	4.58 br s	78.0	4.83 br s	78.4	4.33 br s
5a	53.0	3.70 m	66.8	4.08 2H m	69.1	3.75 m
5b		3.43 m				3.68 m
6a	43.0	2.56 m	41.0	2.59 m	41.0	2.50 m
6b		2.19 m		2.38 m		2.24 m
7	57.9		57.1		57.0
8	134.5		134.1		135.2
9	113.0	6.89 d (7.5)	113.0	7.00 d (7.4)	114.0	7.10 d (7.4)
10	123.7	6.82 t (7.8)	123.8	6.85 t (7.8)	123.8	6.97 t (8.0)
11	117.4	6.74 d (8.1)	117.6	6.76 d (8.1)	112.5	6.92 d (8.2)
12	143.3		143.4		146.2
13	132.7		132.8		133.6
14a	26.5	2.37 dt (14.6, 3.4)	25.9	2.62 m	25.8	2.65 br d (14.5)
14b		1.35 dt (14.6, 2.5)		1.47 br d (14.8)		1.29 br d (14.5)
15	25.8	3.32 m	25.1	3.43 br s	25.7	3.36 br s
16	105.3		105.0		105.6
18	20.5	1.32 d (6.6)	20.8	1.37 d (6.4)	20.8	1.33 d (6.4)
19	69.0	3.82 m	68.4	3.91 m	69.0	3.95 m
20	42.0	2.10 m	44.4	2.13 m	44.8	1.97 m
21a	48.5	3.48 dd (13.5, 11.7)	64.1	4.00 dd (13.4, 6.1)	66.3	3.71 dd (13.6, 6.1)
21b		3.22 dd (13.5, 6.2)		3.81 dd (13.0, 10.6)		3.47 dd (13.4, 9.3)
12-OMe					56.3	3.88 s
CO2Me	168.4		168.1		168.4
CO2 Me	51.7	3.79, s	51.8	3.79 s	51.8	3.78 s

aRecorded at 600 and 150 MHz

Fig. 4

Key ROESY () correlations for alkaloids 4–6

Alkaloid 5 exhibited a molecular ion peak at m/z 372.1676 in its HREIMS spectrum, indicating the molecular formula of C20H24N2O5 with sixteen mass units higher than that of 4. Examination of the 1H NMR spectrum revealed four downfield shifts for the protons H-3 (δH 4.83), H-5 (δH 4.08), and H-21 (δH 4.00 and 3.81), while the 13C NMR data exhibited noticeable downfield shifts involving C-3 (δC 78.0), C-5 (δC 66.8), and C-21 (δC 64.1), when compared to those of 4. These features are characteristic of N(4)-oxides, and this conclusion was further confirmed by the HMBC correlations of δH 4.08 (H2-5), 2.62 (H-14a), and 4.00 (H-21a) with δC 78.0 (d, C-3). The ROESY correlations (Fig. 4) indicated that the relative configuration of 5 was the same as that of 4. Other parts of 5 were identical to those of 4 as secured by detailed analysis of extensive 2D NMR spectral data of 5. Thus, the structure of alkaloid 5 was determined to be as shown in Fig. 1 and named alstoniascholarine P.

Alkaloid 6 was deduced to have the molecular formula C21H26N2O5, as indicated by the observed ion peak at m/z 386.1847 [M]+ in its HREIMS data. The IR, UV and 1D NMR spectra of 6 were similar to those of 5, which suggested that 6 was also a strychnan type indole alkaloid. The 1H and 13C NMR data (Table 3) of 6 exhibited high similarities with those of Nb-demethylalstogustine N-oxide (25), except for the loss of one aromatic proton and the existence of an additional methoxy group (δC 56.3, δH 3.88). By comparing the 1H and 13C NMR spectral data of 6 with those of 25, the C-12 carbon signal was significantly deshielded, while the C-11 and C-13 carbon signals were relatively shielded, suggesting that the methoxy was attached to C-12. This conclusion was further confirmed by the correlation of methoxy protons at δH 3.88 with C-12 (δC 146.2) in the HMBC spectrum. Complete analysis of 2D NMR spectral data of 6 suggested that its other parts were the same to those of 4 and 5. Hence, the structure of alkaloid 6 was assigned as shown in Fig. 1 and named as alstoniascholarine Q.

The known alkaloids picrinine (7) [18], strictamine (8) [19], tubotawine (9) [20], alstolucine D (10) [8], nareline (11) [21], picralinal (12) [18], isoalschomine (13) [18], polyneuridine (14) [22], burnamine (15) [23], echitamidine (16) [24], scholarisine I (17) [25], 19,20-Z-vallesamine (18) [26], 19,20-E-vallesamine (19) [26], scholaricine (20) [27], 19-epi-scholaricine (21) [16], pseudoakuammigine Nb-oxide (22) [28], tubotawine Nb-oxide (23) [29], vallesamine Nb-oxide (24) [16], and Nb-demethylalstogustine N-oxide (25) [30], were identified by comparison of their spectroscopic data with those reported in the literature.

Considering the various bioactivities of the chemical constituents from A. scholaris reported previously [5], six new alkaloids (1–6) were evaluated for their cytotoxic and neurite outgrowth-promoting activities. Unfortunately, the alkaloids 1–6 neither showed cytotoxic activity against five human cancer cell lines (SW-480, SMMC-7721, HL-60, MCF-7, and A-549), nor exhibited neurite outgrowth-promoting activity.

Experimental Section

General Experimental Procedures

Optical rotations were recorded on a JASCO P-1020 polarimeter. UV spectra were measured on a Shimadzu UV-2401PC spectrophotometer. IR spectra were performed on a Bruker FT-IR Tensor-27 infrared spectrophotometer with KBr discs. NMR spectroscopic data were obtained on Bruker AM-400 and DRX-500 spectrometers using TMS as an internal standard. Unless otherwise specified, chemical shifts (δ) were expressed in ppm with reference to the solvent signals. ESIMS spectra were performed on Waters Xevo TQ-S and Bruker HCT/Esquire mass spectrometers; HREIMS and HRESIMS analyses were carried out on Waters AutoSpec Premier P776 and Agilent 6540 UHD Accurate-Mass Q-TOF LC/MS mass spectrometers, respectively. Column chromatography (CC) was conducted on silica gel (200–300 mesh, Qingdao Marine Chemical Co., Ltd., Qingdao, People’s Republic of China), RP-18 gel (20–45 μm, Fuji Silysia Chemical Ltd., Japan), and Sephadex LH-20 (GE Healthcare Bio-sciences AB, Sala, Sweden). Fractions were monitored by TLC (GF 254, Qingdao Marine Chemical Co., Ltd., Qingdao), and spots were visualized by Dragendorff’s reagent.

Plant Material

The inadequately dried leaves of A. scholaris were collected in June 2013 in Pu’er city of Yunnan Province, P. R. China and identified by Dr. Xiao-Dong Luo, Kunming Institute of Botany, Chinese Academy of Sciences. A voucher specimen (No. Luo20130720) has been 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

The inadequately dried leaves (40 kg) of A. scholaris were extracted with MeOH at room temperature, and the solvent was evaporated in vacuo. The extract was dissolved in 0.3 % HCl solution and then partitioned with EtOAc three times. The acidic water-soluble material, subsequently basified with 5 % ammonia solution to pH 9–10, was repeatedly partitioned with EtOAc three times to give the crude alkaloid extract (155 g). The alkaloid extract was subjected to a silica gel column and eluted with CHCl3–MeOH (30:1, 15:1, 10:1, 5:1, 1:1) to afford fractions I–V. Fraction I (40.0 g) was separated by RP-18 CC, eluted with MeOH-H2O (55:1 → 100:1) to give five subfractions (Fr. I–I to Fr. I–V). Fraction I–II (3.8 g) was further separated by silica gel CC, eluted with CHCl3-acetone (10:1 → 5:1) to yield alkaloids 8 (25 mg), 9 (18 mg), 10 (38 mg), and 11 (123 mg). Fraction I–III (16 g) was subjected to RP-18 CC, eluted with MeOH-H2O (75:25 → 95:5) to obtain alkaloids 1 (18 mg), 2 (20 mg), and 7 (9.8 g). Fraction II (10 g) was separated by RP-18 CC, eluted with MeOH-H2O (50:50–80:20) and subsequently purified by Sephadex LH-20 CC to yield alkaloids 12 (17 mg), 13 (12 mg), 14 (10 mg), 15 (15 mg) and 16 (32 mg). Fraction III (35 g) was subjected to RP-18 CC and eluted with MeOH-H2O (40:60 → 65:35) to obtain five fractions (Fr. III–I to Fr. III–V). Fraction III–II (200 mg) was chromatographed repeatedly over silica gel CC, eluted with CHCl3–MeOH (12:1 → 10:1) to give alkaloids 3 (12 mg) and 17 (8 mg). In the same way, alkaloid 20 (6.5 g) was obtained from fraction III–III (8.2 g) by silica gel CC and eluted with CHCl3–MeOH (12:1 → 10:1). Fraction III–IV (1 g) was separated by RP-18 CC, eluted with MeOH-H2O (35:65 → 60:40) to afford alkaloids 4 (7 mg), 18 (45 mg), 19 (18 mg), and 21 (85 mg). Fraction IV (14 g) was separated over RP-18 CC, eluted with MeOH-H2O (25:75 → 60:40) to afford five fractions (Fr. IV–I to Fr. IV–V). Fraction IV–II (120 mg) was subjected to silica gel CC, eluted with CHCl3–MeOH (8:1 → 5:1) to yield alkaloid 22 (15 mg). Fraction IV–III (200 mg) was purified by RP-18 CC, eluted with MeOH-H2O (30:70 → 50:50) to give alkaloids 23 (4 mg) and 24 (6 mg). Fraction IV–IV (300 mg) was subjected to RP-18 CC eluted with MeOH-H2O (25:75 → 50:50) and further purified by silica gel CC eluted with CHCl3–MeOH (6:1 → 4:1) to yield alkaloids 5 (6 mg), 6 (8 mg) and 25 (6 mg).

Alstoniascholarine L (1)

White amorphous powder; –121.1 (c 0.1 MeOH); UV (MeOH) λmax (log ε) 204 (4.45), 240 (3.87), 298 (3.49) nm; IR (KBr) νmax 3440, 2924, 1748, 1630, 1612, 1473, 1090 cm−1; CD (MeOH) λmax (Δε) 202 (+8.1), 239 (–10.8), 302 (–2.1) nm; 1H (400 MHz, CD3OD) and 13C NMR (100 MHz, CD3OD) data, see Table 2; ESIMS m/z 379 [M + Na]+; HREIMS m/z 356.1366 ([M]+ calcd for C19H20N2O5, 356.1372).

Alstoniascholarine M (2)

White amorphous powder; –117.1 (c 0.1 MeOH); UV (MeOH) λmax (log ε) 204 (4.48), 240 (3.90), 298 (3.52) nm; IR (KBr) νmax 3441, 2923, 1748, 1631, 1474, 1241, 1085 cm−1; CD (MeOH) λmax (Δε) 202 (+8.3), 239 (–11.6), 302 (–2.4) nm; 1H (400 MHz, CD3OD) and 13C NMR (100 MHz, CD3OD) data, see Table 2; ESIMS m/z 379 [M + Na]+; HREIMS m/z 356.1358 ([M]+ calcd for C19H20N2O5, 356.1372).

Alstoniascholarine N (3)

White amorphous powder; +2.7 (c 0.1 MeOH); UV (MeOH) λmax (log ε) 204 (4.38), 229 (4.35), 291 (3.63) nm; IR (KBr) νmax 3426, 2976, 1764, 1640, 1511, 1244, 1104 cm−1; 1H (600 MHz, CD3OD) and 13C NMR (150 MHz, CD3OD) data, see Table 2; ESIMS m/z 377 [M + Na]+; HREIMS m/z 354.1577 ([M]+ calcd for C20H22N2O4, 354.1580).

Alstoniascholarine O (4)

White amorphous powder; –348.7 (c 0.1 MeOH); UV (MeOH) λmax (log ε) 213 (4.09), 230 (4.00), 289 (3.61), 337 (4.06) nm; IR (KBr) νmax 3417, 2977, 1678, 1620, 1440, 1260, 1074 cm−1; 1H (600 MHz, CD3OD) and 13C NMR (150 MHz, CD3OD) data, see Table 3; ESIMS m/z 357 [M + H]+; HRESIMS m/z 357.1810 ([M + H]+ calcd for C20H25N2O4, 357.1809).

Alstoniascholarine P (5)

White amorphous powder; –329.3 (c 0.1 MeOH); UV (MeOH) λmax (log ε) 211 (4.12), 287 (3.60), 336 (3.98) nm; IR (KBr) νmax 3431, 2977, 1680, 1621, 1439, 1202, 1059 cm−1; 1H (600 MHz, CD3OD) and 13C NMR (150 MHz, CD3OD) data, see Table 3; ESIMS m/z 373 [M + H]+; HREIMS m/z 372.1676 ([M]+ calcd for C20H24N2O5, 372.1685).

Alstoniascholarine Q (6)

White amorphous powder; –298.7 (c 0.12 MeOH); UV (MeOH) λmax (log ε) 210 (4.21), 287 (3.76), 334 (4.06) nm; IR (KBr) νmax 3438, 2965, 1681, 1615, 1461, 1266, 1060 cm−1; 1H (600 MHz, CD3OD) and 13C NMR (150 MHz, CD3OD) data, see Table 3; ESIMS m/z 387 [M + H]+; HREIMS m/z 386.1847 ([M]+ calcd for C21H26N2O5, 386.1842).

Antitussive Effects Against Ammonia-Induced Cough

The antitussive assay was performed using the technique as described in the literature [31]. ICR (Institute of Cancer Research) mice of either sex (19–24 g) were purchased from Kunming Medical College (license number SYXK 2005-0001). All mice were housed at room temperature (20–25 °C) and constant humidity (40–70 %) under a 12 h light–dark cycle in SPF (Specific Pathogen Free) grade laboratory. The animal study was performed according to the international rules considering animal experiments and the internationally accepted ethical principles for laboratory animal use and care. The mice were divided randomly, of which 12 mice per group. The negative control group of animals was treated with distilled water orally, and the positive control was treated with codeine phosphate, the remaining groups were treated with test samples. Antitussive activity was investigated on a classical mouse cough model induced by ammonia liquor. Briefly, each mouse was placed in a 300 mL special glass chamber and exposed to 40 μL 25 % ammonia solution. The cough frequency produced during 2 min exposure period was counted. In the second assay for alkaloids, cough frequency and latent period of cough were recorded.

Cytotoxic Activity Assay

The following human cancer cell lines were used: SW-480, SMMC-7721, HL-60, MCF-7, and A-549. All cells were cultured in RPMI-1640 or DMEM medium (Hyclone, Logan, UT), supplemented with 10 % fetal bovine serum (Hyclone) at 37 °C in a humidified atmosphere with 5 % CO2. Cell viability was assessed by conducting colorimetric measurements of the amount of insoluble formazan formed in living cells based on the reduction of 3-(4, 5-dimethylthiazol-2-yl)-2, 5-diphenyltetrazolium bromide (MTT) (Sigma, St. Louis, MO) [32]. Briefly, 100 μL of adherent cells was seeded into each well of a 96-well cell culture plate and allowed to adhere for 12 h before drug addition, while suspended cells were seeded just before drug addition, both with an initial density of 1 × 105 cells/mL in 100 μL of medium. Each cell line was exposed to the test compound at various concentrations in triplicate for 48 h, with cisplatin and paclitaxel (Sigma) as positive controls. After the incubation, MTT (100 μg) was added to each well, and the incubation continued for 4 h at 37 °C. The cells were lysed with 100 μL of 20 % SDS-50 % DMF after removal of 100 μL of medium. The optical density of the lysate was measured at 595 nm in a 96-well Microtiter plate reader (Bio-Rad 680).

Neurite Outgrowth-Promoting Activity Assay

The neurotrophic activities of the test compounds were examined according to an assay using PC12 cells as reported elsewhere [33]. Briefly, PC12 cells were maintained in F12 medium supplemented with 12.5 % horse serum (HS), and 2.5 % fetal bovine serum (FBS), and incubated at 5 % CO2 and 37 °C. Test compounds were dissolved in DMSO. For the neurite outgrowth-promoting activity bioassay, PC12 cells were seeded at a density of 5 × 104 cells/mL in 48-well plate coated with poly-l-lysine. After 24 h, the medium was changed to that containing 10 µM of each test compounds plus 5 ng/mL NGF, or various concentrations of NGF (50 ng/mL for the positive control, 5 ng/mL for the negative control). The final concentration of DMSO was 0.05 %, and the same concentration of DMSO was added into the negative control. After 72 h incubation, the neurite outgrowth was assessed under a phase contrast microscope. Neurite processes with a length equal to or greater than the diameter of the neuron cell body were scored as neurite bearing cells. The ratio of the neurite-bearing cells to total cells (with at least 100 cells examined/view area; 5 viewing area/well) was determined and expressed as a percentage.

Supplementary material 1 (PDF 3684 kb)