Phytosteroids and triterpenoids with potent cytotoxicities from the leaves of Chisocheton cumingianus

Abstract

Six phytosteroids and three dammarane-type triterpenoids, namely chisopanoids A–F (1–6) and chisopanones G–I (7–9), together with nine known triterpenoids (10–18) were isolated from the leaves of Chisocheton cumingianus. Their structures were elucidated by extensive spectroscopic analysis, X-ray crystallographic analysis, Mosher's method and Mo₂(OAc)₄-induced electronic circular dichroism (ECD) data. All compounds were evaluated for their cytotoxicities against HepG2, U2OS and MCF-7 human cancer cell lines, as well as inhibitory effects on nitric oxide (NO) production in LPS-activated RAW 264.7 macrophages. Compounds 5 and 6 showed potent cytotoxicities towards MCF-7 with IC₅₀ values of 3.24 ± 1.39 and 8.85 ± 4.73 μM, and were further proved to inhibit the cell proliferation mainly by inducing apoptosis.

---

Introduction

Phytosterols, such as β-sitosterol, stigmasterol, and campesterol, are widespread cholesterol-like chemicals in higher plants, and often undergo oxidation processes to form phytosterol oxidation products (POPs).^1–3 POPs have interested organic chemists and natural products researchers owing to their binding ability with various carrier proteins, which is desirable in most in vivo drug applications.⁴ Previous studies revealed POPs' anticancer potential that varied obviously with different functional groups, especially in ring B.⁵ However, influence of the side chain is not well investigated, mainly due to its limited diversities, especially the oxidated ones which have been reported only once in natural products as far as we know.⁶ As continuous research on Chisocheton cumingianus,^7–9 bioassay-guided isolation on its leaves led to six new cytotoxic POPs with different oxidated side chains, namely chisopanoids A–F (1–6) (Fig. 1). Compounds 5 and 6 showed the most potent cytotoxicities towards MCF-7 with IC₅₀ values of 3.24 ± 1.39 and 8.85 ± 4.73 μM, comparable to clinically used doxorubicin. The induction of apoptosis was then confirmed to be responsible according to the APC Annexin V/7-AAD double-staining assay. In addition, three new and nine known triterpenoids, compounds 7–9 and 10–18, were obtained, and some of which showed inhibitory effects on LPS-activated nitric oxide (NO) production in RAW 264.7 macrophages. Compound 10 was the most active one with an IC₅₀ value of 5.14 ± 2.16 μM, nearly 7 times better than that of the positive control L-NMMA. Herein, we described the isolation, structural elucidation and activity evaluation of all isolates from C. cumingianus.

Fig. 1 Chemical structures of compounds 1–9.

Results and discussion

Chisopanoid A (1) was obtained as colorless crystals. The molecular formula, C₂₉H₄₈O₃, was determined by HRESIMS m/z 467.3493 [M + Na]⁺ (calcd for C₂₉H₄₈NaO₃, 467.3496), requiring six degrees of unsaturation. The IR absorptions implied the presence of hydroxyl (3450 cm⁻¹) and olefinic bond (1639 cm⁻¹). The ¹H NMR spectrum (Table 1) revealed two typical angular methyl groups of steroid (δ_H 0.79 and 1.06, each 3H, s), three secondary methyl groups [δ_H 0.90 (d, 3H, J = 6.8 Hz), 0.95 (d, 3H, J = 6.8 Hz) and 1.10 (d, 3H, J = 6.6 Hz)], four oxygenated protons [δ_H 3.51 (1H, m), 3.52 (1H, m), 3.60 (1H, m) and 3.81 (1H, br s)], and three olefinic protons [δ_H 5.14 (1H, dd, J = 15.2, 9.4 Hz), 5.29 (1H, dd, J = 15.2, 8.5 Hz) and 5.59 (1H, d, J = 5.1 Hz)]. All 29 carbons observed in the ¹³C NMR spectrum (Table 1) were further classified by HSQC experiment as five methyls, nine methylenes (one oxygenated), twelve methines (two oxygenated and three olefinic), and three quaternary carbons (one olefinic). Double bonds took two out of the six indices, and the remaining four suggested the tetracyclic skeleton of 1. The aforementioned data indicated that 1 was a stigmastane-type steroid and was structurally similar to (22E,24S)-stigmast-5,22-dien-3β,7α-diol,¹⁰ except for the presence of an additional hydroxyl group.

Table 1 ¹H NMR (500 MHz) and ¹³C NMR (125 MHz) data of compounds 1–6

Position	1^b		2^b		3^a		4^a		5^b		6^a
	δH (multi, J in Hz)	δC	δH (multi, J in Hz)	δC	δH (multi, J in Hz)	δC	δH (multi, J in Hz)	δC	δH (multi, J in Hz)	δC	δH (multi, J in Hz)	δC
a Measured in CDCl3.b Measured in methanol-d4.
1a	1.90 m	38.1	1.89 m	38.1	1.86 m	37.2	1.86 m	37.2	1.87 m	38.3	1.95 m	36.5
1b	1.22 m		1.20 m		1.12 m		1.13 m		1.05 m		1.21 m
2a	1.84 m	32.2	1.84 m	32.1	1.85 m	31.5	1.84 m	31.6	1.80 m	32.3	1.94 m	31.3
2b	1.57 m		1.54 m		1.51 m		1.51 m		1.50 m		1.63 m
3	3.52 m	72.1	3.51 m	72.1	3.59 m	71.4	3.59 m	71.6	3.41 m	72.2	3.68 m	70.7
4a	2.33 m	42.9	2.33 m	42.9	2.34 m	42.1	2.33 m	42.2	2.25 m	42.6	2.50 m	42.0
4b	2.30 m		2.30 m		2.28 m		2.27 m		2.25 m		2.40 m
5		146.7		146.7		146.3		146.4		144.1		165.1
6	5.59 d (5.1)	125.0	5.57 d (5.2)	125.0	5.60 d (3.9)	124.0	5.60 d (4.4)	124.1	5.28 br s	127.4	5.68 d (1.2)	126.3
7	3.81 br s	65.9	3.79 br s	65.9	3.85 br s	65.4	3.85 br s	65.5	3.72 m	73.8		202.3
8	1.53 m	39.0	1.51 m	39.0	1.46 m	37.6	1.44 m	37.6	1.44 m	41.3	2.24 dd (11.6, 11.5)	45.6
9	1.37 m	43.5	1.36 m	43.4	1.22 m	42.4	1.22 m	42.5	1.04 m	50.1	1.56 m	50.1
10		38.5		38.5		37.5		37.7		37.4		38.4
11a	1.61 m	21.8	1.64 m	21.8	1.53 m	20.8	1.52 m	21.0	1.60 m	22.3	1.58 m	21.4
11b	1.61 m		1.54 m		1.53 m		1.52 m		1.51 m		1.58 m
12a	2.05 dt (12.6, 3.2)	40.6	2.05 dt (12.6, 3.2)	40.6	2.00 dt (12.6, 3.3)	39.3	1.99 dt (12.6, 3.2)	39.4	2.05 dt (12.6, 3.3)	41.1	2.03 dt (12.9, 3.3)	38.9
12b	1.24 m		1.24 m		1.17 m		1.17 m		1.17 m		1.15 m
13		43.2		43.2		42.2		42.3		44.1		43.3
14	1.53 m	50.8	1.52 m	50.7	1.46 m	49.5	1.46 m	49.6	1.15 m	57.8	1.51 m	50.2
15a	1.81 m	25.1	1.78 m	25.2	1.72 m	24.4	1.72 m	24.5	1.87 m	27.3	2.41 m	26.5
15b	1.16 m		1.12 m		1.13 m		1.14 m		1.43 m		1.35 m
16a	1.82 m	30.2	1.80 m	30.2	1.90 m	28.4	1.91 m	28.5	1.88 m	29.6	1.91 m	28.7
16b	1.36 m		1.35 m		1.29 m		1.29 m		1.30 m		1.29 m
17	1.27 m	57.3	1.27 m	57.4	1.18 m	55.8	1.18 m	55.7	1.12 m	56.9	1.11 m	54.9
18	0.79 s	12.3	0.78 s	12.3	0.69 s	11.7	0.68 s	11.8	0.73 s	12.4	0.68 s	12.1
19	1.06 s	18.7	1.04 s	18.6	1.00 s	18.3	0.99 s	18.4	1.07 s	19.5	1.20 s	17.5
20	2.13 m	41.8	2.15 m	41.9	1.38 m	36.1	1.44 m	36.6	1.39 m	37.6	1.38 m	36.2
21	1.10 d (6.6)	21.7	1.10 d (6.7)	21.7	0.93 d (6.5)	18.9	0.97 d (6.5)	19.0	0.96 d (6.5)	19.4	0.92 d (6.5)	19.1
22a	5.29 dd (15.2, 8.5)	140.0	5.31 dd (15.2, 8.8)	141.4	1.19 m	34.0	1.44 m	36.6	1.08 m	35.0	1.20 m	34.1
22b					1.05 m		1.13 m		1.08 m		1.05 m
23a	5.14 dd (15.2, 9.4)	130.2	5.11 dd (15.2, 9.5)	129.4	1.19 m	27.0	1.45 m	22.8	1.24 m	28.1	1.20 m	27.2
23b					1.19 m		1.15 m		1.24 m		1.20 m
24	1.87 m	47.0	1.74 m	58.1	1.16 m	40.9	0.98 m	51.6	1.17 m	42.1	1.18 m	40.9
25	1.60 m	33.5		73.3	1.69 m	29.9	1.78 m	29.1	1.69 m	31.0	1.69 m	29.9
26	0.95 d (6.8)	21.3	1.18 s	29.6	0.85 d (6.3)	19.3	0.94 d (6.9)	21.2	0.87 d (6.8)	19.5	0.84 d (7.0)	19.4
27	0.90 d (6.8)	19.4	1.12 s	25.8	0.84 d (6.3)	19.1	0.88 d (6.8)	19.9	0.86 d (6.8)	19.7	0.83 d (6.8)	19.2
28a	1.71 m	36.6	1.75 m	22.5	1.56 m	34.2	3.92 m	69.4	1.55 m	34.9	1.57 m	34.3
28b	1.47 m		1.23 m		1.40				1.41 m		1.40 m
29a	3.60 m	61.7	0.87 t (7.3)	13.0	3.66 m	62.1	1.18 d (6.5)	22.2	3.56 m	62.0	3.65 m	62.2
29b	3.51 m				3.64 m						3.65 m

---

HMBC correlations (Fig. 2) traced from the methyl protons to their neighboring carbons enabled the establishment of stigmastane skeleton. Two double bonds were placed across C-5/C-6 and C-22/C-23 due to the HMBC correlations from H-6 to C-7/C-8/C-10/C-4, from H-22 to C-20/C-21/C-23/C-24, and from H-23 to C-20/C-22/C-24/C-25/C-28. Besides, the olefinic bond between C-22 and C-23 was assigned to be an E geometry due to its large vicinal coupling (J = 15.2 Hz). In combination with the chemical shift, the cross peaks from H-1a/H-4 to C-3, from H-6 to C-7, and from H-29a to C-24/C-28 implied the locations of three hydroxy groups at C-3, C-7 and C-29 respectively. Thus, the gross structure of 1 was established as shown (Fig. 1). The relative configuration of 1 was elucidated mainly from the analysis of its ROESY spectrum (Fig. 2). Correlations of Me-19/H-1a, Me-18/Me-19, Me-18/H-12a, and Me-18/H-8 revealed that they were cofacial and assigned as β-oriented. In turn, the ROESY correlations of H-3/H-1b, H-12b/H-9, H-12b/H-14, H-12b/H-17 and H-12b/Me-21 suggested that they are cofacial and α-oriented. As for the 7-OH, since chemical shift of H-6 (δ_H 5.60/5.29) and C-7 (δ_C 65.5/73.3) are highly diagnostic between 7α and 7β-hydroxy-5-ene sterols,¹¹ 1 was obviously substituted with a 7α-OH (Table 1). To further confirm the structure and establish the absolute configuration, single crystals of 1 were obtained and subjected to an X-ray diffraction experiment using a mirror Cu Kα radiation.¹² As shown in Fig. 3, the structure of 1 was further proved and certified to be 3S, 7S, 8S, 9S, 10R, 13R, 14S, 17R, 20R and 24S configurations with a Flack parameter value of −0.07(6). Hence, 1 was determined as (3S,7S,20R,22E,24S)-stigmast-5,22-dien-3,7,29-triol.

Fig. 2 Key HMBC and ROESY correlations of compound 1.

Fig. 3 X-Ray crystal structure of 1.

Chisopanoid B (2) closely resembled 1 except for the location of the hydroxy on the side chain. The singlets of Me-26, Me-27 [δ_H 1.18 (3H, s), 1.12 (3H, s)] in 2 which were doublets in 1, along with the appearance of a triplet methyl signal at δ_H 0.87 (3H, t, J = 7.3 Hz), suggested that the hydroxy group shifted from C-29 to C-25. The HMBC correlations from Me-26 and Me-27 to C-25 further proved the supposed structure. As a result, 2 was assigned as (3S,7S,20R,22E,24S)-stigmast-5,22-dien-3,7,25-triol.

Chisopanoid C (3) was isolated as a white powder. Its HRESIMS showed a positive ion at m/z 469.3655 [M + Na]⁺ (calcd for C₂₉H₅₀NaO₃, 469.3652), corresponding to a molecular formula of C₂₉H₅₀O₃, 2 mass units more than that of 1. NMR data comparison between 3 and 1 showed their identical tetracyclic moieties but different side chains. In accordance with the molecular formula and HMBC correlations, the Δ²²⁽²³⁾ olefinic signal in 1 was transformed to single bond, which established the sitostane skeleton of 3. ROESY experiments indicated that the relative configuration of 3 was similar to that of 1. Chisopanoid D (4), an isomer of 3, shared the same skeleton with mild difference in the side chain. The hydroxy group at C-29 in 3 transferred to C-28 in 4, resulting in the emergence of an additional methyl group δ_H 1.18 (d, J = 6.5 Hz) and the vanishment of an oxygenated proton. The same relative configuration of 3β-OH, 7α-OH and 21α-Me with above-mentioned compounds was advised from the ROESY experiment. The absolute configuration was determined by Mosher's method.¹³ Treatment of 4 with (R)-(−)- and (S)-(+)-α-methoxy-α-(trifluoromethyl)-phenylacetyl (MTPA-Cl) gave the (S)- and (R)-MTPA esters 4a and 4b, respectively. The ¹H NMR signals of the two MTPA esters were assigned on the basis of their ¹H–¹H COSY spectra, and the Δδ_H(S–R) values were then calculated as shown in Fig. 4. Therefore, the absolute configuration of 4 was determined as 3S and 28R, and thus 4 was established as (3S,7S,20R,28R)-stigmast-5-ene-3,7,28-triol. As 3 owned semblable NMR data and electronic Cotton effects (see ESI Fig. S22 and S29†) to 4, it had identical absolute configuration and was then determined to be (3S,7S,20R)-stigmast-5-ene-3,7,29-triol.

Fig. 4 Chemical shift differences (Δδ_H(S–R)) between (S)-MTPA and (R)-MTPA esters of 4.

Chisopanoid E (5) was an analogue of 3, differed only in the configuration of 7-OH. The characteristic signal of H-6 (δ_H 5.28, br s) and C-7 (δ_C 73.8) demonstrated that 5 was substituted with a 7β-hydroxy group,¹¹ accordingly 7R configuration. Thus, 5 was named as (3S,7R,20R)-stigmast-5-ene-3,7,29-triol. Chisopanoid F (6), obtained as a colorless gum, possessing the molecular formula of C₂₉H₄₈O₃ by HRESIMS m/z 467.3498 [M + Na]⁺ (calcd for C₂₉H₄₈NaO₃, 467.3496), 2 mass units less than that of 3. Analysis of its NMR data revealed that 6 was structurally similar to 3. But the existence of α,β-unsaturated ketone moiety (λ_max 238 nm in UV) indicated the oxidation of 7-OH in 6, causing the downfield-shifted carbon resonances at δ_C 202.3 (Δδ +136.9, C-7), δ_C 165.1 (Δδ +18.8, C-5), 126.3 (Δδ +2.3, C-6), 45.6 (Δδ +8.0, C-8), 50.1 (Δδ +7.7, C-9) and 26.5 (Δδ +2.1, C-15). This deduction was further confirmed by the HMBC correlations from H-8/H-14 to C-7. Additionally, 6 possessed the same configuration as 3 and was determined to be (3S,20R)-3,29-dihyodroxy-stigmast-5-ene-7-one.

Chisopanone F (7) was obtained as a white powder. HRESIMS gave a molecular formula of C₃₀H₅₀O₄, deduced from the positive ion peak at m/z 497.3603 [M + Na]⁺ (calcd for C₃₀H₅₀NaO₄, 497.3601), requiring six indices of unsaturation. The ¹H NMR spectrum (Table 2) displayed seven tertiary methyl groups (δ_H 0.71, 0.83, 0.97, 1.03, 1.14, 1.17 and 1.17, each 3H, s), two oxygenated protons [3.27 (1H, m) and δ_H 3.97 (1H, s)], and a terminal double bond (δ_H 4.74 and 4.76, each 1H, br s). In ¹³C NMR spectrum (Table 2), 30 well resolved carbon signals were found, which were classified with the aid of HSQC experiment as seven methyls, ten methylenes, six methines (two oxygenated) and seven quaternary carbons (one oxygenated and one carbonyl). All evidence advised a dammarane-type triterpene skeleton, structurally similar to 24,25-dihydroxydammar-20-en-3-one.¹⁴ The carbonyl group moved upfield to δ_C 212.9, and a sharp oxygenated proton at δ_H 3.97 (1H, s) emerged in 7, corresponding to those of ring A in viburnudienones H₁.¹⁵ HMBC correlations (Fig. 5) from H-1a/H-1b/H-3 to C-2, and from H-1a/Me-28/Me-29 to C-3 further confirmed the ring A moiety with carbonylated C-2 and hydroxylated C-3. ROESY experiment of 7 gave the relative configuration of the dammarane-type tetracyclic core (Fig. 5), and the correlations of Me-19/H-1a, H-3/H-1b, H-3/H-5, H-3/Me-28 indicated the α-orientation for H-3. The absolute configuration of C-24 was assigned by using the in situ Mo₂(OAc)₄-induced electronic circular dichroism (ECD) method.¹⁶ According to the empirical rule proposed by Snatzke,^17,18 the well induced ECD spectrum (see ESI Fig. S53†) with positive Cotton effect at 305 nm permitted the assignment of 24S configuration. In summary, 7 was established as (24S)-3β,24,25-trihydroxydammar-20-en-2-one.

Table 2 ¹H NMR (500 MHz) and ¹³C NMR (125 MHz) data of compounds 7–9

Position	7^b		8^b		9^a
	δH (multi, J in Hz)	δC	δH (multi, J in Hz)	δC	δH (multi, J in Hz)	δC
a Measured in CDCl3.b Measured in methanol-d4.
1a	2.42 d (12.0)	55.1	2.41 d (12.0)	55.1	1.41 m	33.8
1b	2.22 d (12.0)		2.22 d (12.0)		1.31 m
2a		212.9		212.9	1.95 m	25.6
2b					1.55 m
3	3.97 s	83.9	3.97 s	84.0	3.40 br s	76.4
4		46.5		46.5		37.8
5	1.57 m	56.5	1.57 m	56.4	1.27 m	49.7
6a	1.69 m	19.7	1.69 m	19.7	1.43 m	18.4
6b	1.58 m		1.58 m		1.43 m
7a	1.76 m	36.2	1.81 m	36.0	1.58 m	35.3
7b	1.38 m		1.36 m		1.27 m
8		42.1		42.0		40.8
9	1.75 m	50.7	1.77 m	51.7	1.45 m	50.6
10		44.8		44.8		37.5
11a	1.46 m	22.7	1.43 m	22.9	1.56 m	21.5
11b	1.31 m		1.28 m		1.27 m
12a	1.45 m	30.1	1.92 m	28.2	1.84 m	27.7
12b	1.17 m		1.27 m		1.25 m
13	1.77 m	46.3	1.65 m	44.1	1.66 m	42.7
14		46.9		51.2		50.7
15a	1.68 m	32.5	1.53 m	32.5	1.47 m	31.2
15b	1.17 m		1.11 m		1.11 m
16a	1.62 m	26.0	1.52 m	26.8	1.76 m	25.1
16b	1.16 m		1.10 m		1.48 m
17	1.75 m	51.9	1.84 m	50.9	1.75 m	50.9
18	1.03 s	15.8	1.00 s	15.6	0.97 s	15.7
19	0.83 s	17.6	0.82 s	17.5	0.87 s	16.2
20		153.9		87.7		75.6
21a	4.76 br s	108.3	1.13 s	23.6	1.17 s	26.5
21b	4.74 br s
22a	2.30 m	32.4	1.80 m	36.4	2.44 dd (13.8, 6.9)	43.8
22b	2.30 m		1.67 m		2.35 dd (13.8, 8.5)
23a	1.98 m	31.4	1.79 m	27.4	6.91 ddd (16.0, 8.5, 6.9)	144.6
23b	1.98 m		1.52 m
24	3.27 m	79.4	3.75 dd (7.2, 7.2)	84.8	6.12 d (16.0)	134.0
25		73.8		72.8		198.4
26	1.14 s	25.1	1.15 s	25.4	2.27 s	27.1
27	1.17 s	25.5	1.13 s	26.2
28	1.17 s	29.5	1.16 s	29.5	0.84 s	22.3
29	0.71 s	16.9	0.71 s	16.9	0.95 s	28.5
30	0.97 s	16.4	0.96 s	16.9	0.90 s	16.7

---

Fig. 5 Key HMBC and ROESY correlations of compound 7.

Chisopanone H (8), a white amorphous powder, was an isomer of 7. Due to the NMR data analysis, 8 was also a dammarane-type triterpene, and differed from 7 in the side chain. The disappearance of the terminal double bond signals but unchanged degrees of unsaturation suggested a (20,24)-epoxydammarane triterpene structure, which allowed the downfield-shifted carbon resonance at δ_C 84.8 (Δδ +5.4, C-24) and emergence of the additional oxygenated carbon resonance at δ_C 87.7 (C-20). The HMBC correlations from Me-21 to C-17/C-20/C-22, and from H-24 to C-25/C-27 further verified the epoxy ring. The ROESY spectrum proved that 8 possessed the identical 3β-OH as that of 7. Moreover, correlations between Me-18/H-13, H-13/Me-21, Me-21/H-24, along with the characteristic carbon resonances of the side chain, supported the proposal for 20S,24R stereochemistry on the basis of the molecular model.¹⁹ Thus, 8 was determined to be (20S,24R)-3β,25-dihydroxy-20,24-epoxydammaran-2-one.

Chisopanone I (9) was obtained as a white powder and its molecular formula was determined to be C₂₉H₄₈O₃ by HRESIMS ion at m/z 467.3494 [M + Na]⁺ (calcd for C₂₉H₄₈NaO₃, 467.3496), suggesting six degrees of unsaturation. The ¹H NMR spectroscopic data of 9 (Table 2) displayed two olefinic protons (δ_H 6.12 and 6.91) in a trans relationship (J = 16.0 Hz), and seven methyl signals (δ_H 0.84, 0.87, 0.90, 0.95, 0.97, 1.17, and 2.27, each 3H, s). The ¹³C NMR spectrum (Table 2) with the help of HSQC experiment revealed 29 carbon resonances including one double bond, seven methyls, nine methylenes, five methines and six quaternary carbons (one oxygenated and one carbonyl), suggesting that 9 was identical to 27-demethyl-20(S)-dammar-23-ene-20-ol-3,25-dione,²⁰ a 27-demethyl dammarane-type nortriterpene. Detailed analysis of NMR data showed the same framework except for the reduced hydroxyl group at C-3 in 9, which was further confirmed by HMBC correlations. Furthermore, ROESY correlations of H-3/H-5 and H-5/Me-29 indicated a β-orientation for H-3. Thus, 9 was established as 27-demethyl-(3R,20S)-dammar-23E-ene-3,20-diol-25-one.

By comparison of the NMR and MS spectra with the published data, nine known compounds were identified as 27-demethyl-20(S)-dammar-23-ene-20-ol-3,25-dione (10),²⁰ (20S,25)-dihydroxy-dammar-23-en-3-one (11),²¹ isofouquierone peroxide (12),²² (20S,24S)-dihydroxydammar-25-en-3-one (13),²¹ (24R,25)-dihydroxydammar-20-en-3-one (14),²³ aglaiabbreviatin E (15),²⁴ (20S)-dammar-24-ene-3β,20,26-triol (16),²⁵ 24-(E)-3-oxo-dammara-20,24-dien-26-ol (17),²⁶ 20S-dammar-24-ene-3β,20-diol (18),²⁵ respectively.

All compounds were evaluated for their cytotoxicities against HepG2, U2OS and MCF-7 cancer cell lines with doxorubicin as the positive control. The results were presented in Table 3. Among them, 5 and 6 showed most potent cytotoxic activities against MCF-7 with IC₅₀ values of 3.24 ± 1.39 and 8.85 ± 4.73 μM, which also behaved moderate cytotoxicities towards HepG2 and U2OS cell lines. To clarify the mechanism that induced MCF-7 cells death, the APC Annexin V and 7-AAD costaining flow cytometric analysis was performed. The results demonstrated that 5 triggered massive apoptosis in a dose-dependent manner (Fig. 6), suggesting that they might inhibit the proliferation of MCF-7 cells by inducing apoptosis.

Table 3 The cytotoxic activities of the isolates from C. cumingianus^a

Compounds	HepG-2	U20S	MCF-7
a Results were expressed as IC50 values in μM (n = 3).b Positive control.
1	>50	>50	40.28 ± 6.15
2	>50	10.01 ± 3.25	17.95 ± 3.61
3	28.78 ± 4.92	27.84 ± 5.89	17.98 ± 2.23
4	>50	>50	27.85 ± 2.66
5	14.47 ± 2.77	6.89 ± 1.57	3.24 ± 1.39
6	16.84 ± 0.75	13.42 ± 2.88	8.85 ± 4.73
7	>50	>50	>50
8	>50	>50	>50
9	>50	>50	>50
11	>50	11.67 ± 3.66	26.35 ± 2.01
Doxorubicin^b	3.05 ± 0.32	1.32 ± 0.19	2.05 ± 1.40

---

Fig. 6 MCF-7 cells were incubated with the indicated concentrations of chisopanoid E (5) for 48 h and stained with Annexin V and 7-AAD. Quantitation of the staining data presented as the percentages of cells undergoing apoptosis. Data represents as mean ± SD from three independent experiments. *P < 0.05 and **P < 0.01, compared to control.

To all three human tumor cells, sitostane-type steroids with 7β-hydroxy (5) displayed most potent cytotoxicities, followed by 7-keto group (6) and 7α-hydroxy (3) successively, implying the importance of the substituent attached to C-7. As reported, cholestane-type steroids linked with 7β-hydroxy exhibited better activity than 7α-hydroxy on U937 cells,²⁷ which was consistent with our findings with sitostane-type steroids. On the other hand, when compared with 3, the presence of the 22,23-double bond in the analogue 1 resulted in noticeable decrease of the activities to all three cell lines, in accordance with the report that β-sitosterol oxides displayed better cytotoxicities than stigmasterol oxides.²⁸ Besides, 3β-hydroxy-stigmast-5-ene-7-one was reported inactive to HepG2 cells, but compound 6 only with one more hydroxy group on C-29 demonstrated strong cytotoxicities, suggesting that branched chain oxidation might affect its biological activity.

The individual compounds were also evaluated for their inhibitory effects on NO production in LPS-activated RAW 264.7 macrophages with L-NMMA as the positive control. Notably, triterpenoid 10 showed significant activities against RAW 264.7 cells with IC₅₀ value of 5.14 ± 2.16 μM, nearly 7 times better than that of the positive control L-NMMA (IC₅₀ = 37.46 ± 2.75 μM). As well, triterpenoid 7 and 17 displayed moderate activities with IC₅₀ values of 30.85 ± 2.75 and 22.36 ± 3.70 μM respectively.

Conclusions

Bioassay-guided isolation led to six new POPs with different oxidated side chains from the leaves of C. cumingianus. As expected, they showed different cytotoxicities against HepG2, U2OS and MCF-7 human cancer cell lines. The most effective one, (3S,7R,20R)-stigmast-5-ene-3,7,29-triol, inhibited the proliferation of MCF-7 cells through apoptosis pathway with dose-dependent feature, on a par with clinical used doxorubicin. For cytotoxicities of POPs, the side chains might be suitable modification sites, for that the hydroxylation did not decrease activities of those isolates but increased for compound 6 obviously, when compared to their analogues.^3,29 In addition, although protolimonoids were widely existed in the twigs of C. cumingianus,^7–9 none was found from the leaves by far, implying different chemical compositions of organs as other plants from meliaceae family.^21,30

Experimental

General experimental procedures

Optical rotations were determined with a JASCO P-1020 polarimeter. ECD spectra were carried out on a JASCO-810 spectropolarimeter. UV spectra were obtained on a Shimadzu UV-2501 spectrophotometer. IR (KBr disks) spectra were measured by a Bruker Tensor 27 spectrometer. NMR spectra were recorded on a Bruker Avance III NMR (¹H 500 MHz, ¹³C 125 MHz) instrument at 300 K, with TMS as internal standard. HRESIMS was carried out on an Agilent UPLC-Q-TOF (6520B) mass spectrometer. Melting point was measured on an XT-4 micromelting point instrument. Preparative HPLC was carried out using an SHMADZU LC-6A series instrument with a Shim-park RP-C18 column (200 × 20 mm) and a SHMADZU SPD-20A detector. Silica gel (Qingdao Haiyang Chemical Co., Ltd., Qingdao, China), RP-C18 (40–63 μm, Fuji, Japan) and Sephadex LH-20 (Pharmacia, USA) were used for column chromatography. All solvents used were analytical grade (Jiangsu Hanbang Science and Technology. Co., Ltd.).

Plant material

The leaves of C. cumingianus were collected from Xishuangbanna, Yunnan Province, China, in March 2014. The botanical identification was made by Professor Min-Jian Qin, Department of Medicinal Plants, China Pharmaceutical University. A voucher specimen (no. 201403-CP) is deposited in the Department of Natural Medicinal Chemistry, China Pharmaceutical University.

Extraction and isolation

The air-dried and crushed leaves (1.0 kg) were refluxed with 95% EtOH three times and concentrated in vacuo to give a crude extract (53.5 g), which was suspended in 1 L water and then extracted with petroleum ether to remove the fatty components completely. The defatted crude extracts were further partitioned with CH₂Cl₂ to give the CH₂Cl₂-soluble fraction (31.5 g). Since the CH₂Cl₂ part showed cytotoxic activity in vitro, it was chromatographed on silica gel column eluted with CH₂Cl₂/MeOH in a step manner (100 : 1 to 0 : 100, v/v) to afford six fractions (A–F). Among them, fraction D (6.0 g) behaved strongest cytotoxicity, which was submitted to a MCI gel column eluted with MeOH/H₂O (30 : 70 to 100 : 0, v/v) to give four subfractions (D1–D4). Fraction D3 (1.2 g) was subjected to a reversed-phase C18 column (MeOH/H₂O, 40 : 60 to 100 : 0, v/v) to obtain five subfractions (D3A–D3E). Further purification of fraction D3D by preparative HPLC (MeCN/H₂O, 75 : 25, v/v) generated 1 (11.2 mg), 3 (20.5 mg), 4 (6.2 mg) and 16 (4.6 mg). Taking the similar method, fraction D3E yielded compounds 2 (2.3 mg) and 5 (5.3 mg). In addition, fraction C (7.0 g) was applied to a MCI gel column eluted with MeOH/H₂O (30 : 70 to 100 : 0, v/v) to afford six subfractions (C1–C6). Fraction C4 (2.7 g) was chromatographed on a silica gel column using petroleum ether–acetone (5 : 1 to 1 : 1, v/v), and further purified by preparative HPLC (MeCN/H₂O, 70 : 30, v/v) to obtain 7 (5.8 mg), 8 (4.2 mg), 9 (2.5 mg), and 11 (14.9 mg). Fraction C5 (1.5 g) was isolated by a silica gel column, eluted with petroleum ether–acetone (4 : 1), to give 13 (11.3 mg) and 6 (3.4 mg). Separation of fraction B (12.2 g) via silica gel and Sephadex LH-20 gel columns successively produced six subfractions (B3A–B3F). Fraction B3D (1.5 g) was subjected to an RP-C18 column using a gradient of MeOH–H₂O (50 : 50 to 100 : 0, v/v), to give three subfractions (B3D1–B3D3). Fraction B3D2 (86.2 mg) was purified by preparative HPLC (MeOH/H₂O 90 : 10, v/v) to yield 14 (18.0 mg), 10 (2.9 mg), and 12 (7.1 mg). Using the same purification procedures, fraction B3B (76.1 mg) yielded 15 (4.2 mg), 17 (14.5 mg) and 18 (4.3 mg).

Chisopanoid A (1). Colorless crystals (MeOH–H₂O = 5 : 1); mp 187–189 °C; [α]²⁵_D −79.8 (c 0.12, MeOH); ECD (MeCN) λ (Δε) 197 (−14.819), 221 (+0.246); UV (MeOH) λ_max (log ε) 204 (3.48) nm; IR (KBr) v_max 3450, 2969, 2865, 2316, 1639, 1465, 1384, 1251, 1076 cm⁻¹; ¹H NMR and ¹³C NMR data, see Table 1; HRESIMS m/z 467.3493 [M + Na]⁺ (calcd for C₂₉H₄₈NaO₃, 467.3496).

Chisopanoid B (2). White amorphous powder; [α]²⁵_D −62.2 (c 0.10, MeOH); ECD (MeCN) λ (Δε) 194 (−20.333), 221 (+0.294); UV (MeOH) λ_max (log ε) 204 (3.45) nm; IR (KBr) v_max 3457, 2973, 2935, 2867, 2313, 1640, 1465, 1386, 1068 cm⁻¹; ¹H NMR and ¹³C NMR data, see Table 1; HRESIMS m/z 467.3498 [M + Na]⁺ (calcd for C₂₉H₄₈NaO₃, 467.3496).

Chisopanoid C (3). White amorphous powder; [α]²⁵_D −63.0 (c 0.29, MeOH); ECD (MeCN) λ (Δε) 198 (−18.095), 221 (+0.221); UV (MeOH) λ_max (log ε) 204 (3.61), 239 (2.21) nm; IR (KBr) v_max 3444, 2937, 2870, 2320, 1641, 1464, 1384, 1243, 1055 cm⁻¹; ¹H NMR and ¹³C NMR data, see Table 1; HRESIMS m/z 469.3655 [M + Na]⁺ (calcd for C₂₉H₅₀NaO₃, 469.3652).

Chisopanoid D (4). White amorphous powder; [α]²⁵_D −60.6 (c 0.20, MeOH); ECD (MeCN) λ (Δε) 198 (−10.268), 221 (+0.143); UV (MeOH) λ_max (log ε) 204 (3.49) nm; IR (KBr) v_max 3455, 2967, 2856, 2320, 1640, 1461, 1384, 1262, 1077 cm⁻¹; ¹H NMR and ¹³C NMR data, see Table 1; HRESIMS m/z 469.3650 [M + Na]⁺ (calcd for C₂₉H₅₀NaO₃, 469.3652).

Chisopanoid E (5). White amorphous powder; [α]²⁵_D +21.3 (c 0.20, MeOH); UV (MeOH) λ_max (log ε) 205 (3.53) nm; IR (KBr) v_max 3446, 2935, 2868, 2320, 1642, 1465, 1383, 1058 cm⁻¹; ¹H NMR and ¹³C NMR data, see Table 1; HRESIMS m/z 467.3498 [M + Na]⁺ (calcd for C₂₉H₄₈NaO₃, 467.3496).

Chisopanoid F (6). Colorless gum; [α]²⁵_D −86.7 (c 0.23, MeOH); UV (MeOH) λ_max (log ε) 199 (3.42), 238 (3.72), 278 (3.60) nm; IR (KBr) v_max 3451, 2966, 2867, 2312, 1641, 1465, 1384, 1077 cm⁻¹; ¹H NMR and ¹³C NMR data, see Table 1; HRESIMS m/z 467.3498 [M + Na]⁺ (calcd for C₂₉H₄₈NaO₃, 467.3496).

Chisopanone G (7). White amorphous powder; [α]²⁵_D +49.6 (c 0.22, MeOH); UV (MeOH) λ_max (log ε) 204 (3.30), 270 (2.13) nm; IR (KBr) v_max 3444, 2969, 2854, 2320, 1639, 1463, 1385, 1257, 1079 cm⁻¹; ¹H NMR and ¹³C NMR data, see Table 2; HRESIMS m/z 497.3603 [M + Na]⁺ (calcd for C₃₀H₅₀NaO₄, 497.3601).

Chisopanone H (8). White amorphous powder; [α]²⁵_D +31.2 (c 0.12, MeOH); UV (MeOH) λ_max (log ε) 204 (3.17), 269 (2.28) nm; IR (KBr) v_max 3455, 2971, 2863, 2349, 1642, 1467, 1386, 1082 cm⁻¹; ¹H NMR and ¹³C NMR data, see Table 2; HRESIMS m/z 497.3605 [M + Na]⁺ (calcd for C₃₀H₅₀NaO₄, 497.3601).

Chisopanone I (9). White amorphous powder; [α]²⁵_D −32.8 (c 0.14, MeOH); UV (MeOH) λ_max (log ε) 203 (3.14), 219 (3.09), 291 (2.40) nm; IR (KBr) v_max 3452, 2922, 2854, 2320, 1639, 1384, 1260, 1077 cm⁻¹; ¹H NMR and ¹³C NMR data, see Table 2; HRESIMS m/z 467.3494 [M + Na]⁺ (calcd for C₂₉H₄₈NaO₃, 467.3496).

X-Ray crystallographic study of 1

Colorless crystals of 1 were obtained from MeOH–H₂O (5 : 1). Crystal data were obtained on a Bruker Smart-1000 CCD with a graphite monochromator with Cu Kα radiation (λ = 1.54184 Å) at 290(2) K. The structure was solved by direct methods using the SHELXS-97 (ref. 31) and expanded using difference Fourier techniques, refined with the SHELXL-97.³² Crystallographic data for the structure of 1 has been deposited in the Cambridge Crystallographic Data Centre with the deposition number of CCDC 1433008.†

Crystal data of 1. C₂₉H₄₈O₃ (M = 444.67); monoclinic crystal (0.32 × 0.25 × 0.22 mm³); space group P1; unit cell dimensions a = 8.3207(2) Å, b = 12.7611(3) Å, c = 15.3831(5) Å, β = 79.918(2)°, V = 1458.86(8) ų; Z = 2; ρ_calcd = 1.012 g cm⁻³; μ (Cu Kα) = 0.487 mm⁻¹; 22 972 reflections measured (10.836 ≤ 2Θ ≤ 140.112); 9934 unique (R_int = 0.0155, R_sigma = 0.0164) which were used in all calculations; the final refinement gave R₁ = 0.0501 (>2sigma(I)) and wR₂ = 0.1612 (all data); flack parameter = −0.07 (6).

Preparation of the (S)- and (R)-MTPA esters derivatives of 4

Dimethylaminopyridine 2 mg and (R)-(−)-α-methoxy-α-(trifluoromethyl)-phenylacetyl (MTPA-Cl) 3 μL were added to a solution of 4 (1 mg) in dried pyridine (100 μL), and then the mixture was allowed to stand overnight at 30 °C. MeOH (2 mL) was added to quench the reaction and finally purified by preparative HPLC using MeOH–H₂O (95 : 5, v/v) to give pure (S)-MTPA ester 4a (0.9 mg). The (R)-MTPA ester 4b was prepared using the method described above.

4a. ¹H NMR (CDCl₃, 500 MHz): δ_H 0.62 (3H, s, Me-18), 1.01 (3H, s, Me-19), 0.85 (3H, d, J = 6.6 Hz, Me-21), 0.84 (3H, d, J = 6.0 Hz, Me-26), 0.78 (3H, d, J = 6.8 Hz, Me-27), 1.34 (3H, d, J = 6.5 Hz, Me-29), 1.93 (1H, m, H-1a), 1.65 (1H, m, H-1b), 1.93 (1H, m, H-2a), 1.64 (1H, m, H-2b), 4.87 (1H, m, H-3), 2.51 (2H, m, H-4), 5.85 (1H, d, J = 4.3 Hz, H-6).

4b. ¹H NMR (CDCl₃, 500 MHz): δ_H 0.64 (3H, s, Me-18), 0.98 (3H, s, Me-19), 0.90 (3H, d, J = 6.9 Hz, Me-21), 0.88 (3H, d, J = 5.6 Hz, Me-26), 0.87 (3H, d, J = 6.5 Hz, Me-27), 1.27 (3H, d, J = 6.5 Hz, Me-29), 1.96 (1H, m, H-1a), 1.72 (1H, m, H-1b), 1.93 (1H, m, H-2a), 1.69 (1H, m, H-2b), 4.77 (1H, m, H-3), 2.37 (2H, m, H-4), 5.78 (1H, d, J = 4.9 Hz, H-6).

In vitro cytotoxicity assay

All compounds were evaluated for their cytotoxicities against HepG2, U2OS, and MCF-7 by MTT assay.³³ Cells were plated in 96-well culture plates (5 × 10³ cells per well). After incubation over night, the cells were treated with different concentrations of each compound for 48 h. DMSO (0.1%) was used as a vehicle. MTT (5 mg mL⁻¹) was dissolved in PBS and filter sterilized, then 20 μL of the prepared solution was added to each well and cells were incubated until a purple precipitate was visible. The formed formazan crystals were dissolved in DMSO (150 μL per well) by constant shaking for 10 min. The absorbance was measured on an ELISA reader (SpectraMax Plus 384, Molecular Devices, Sunnyvale, CA) at a test wavelength of 570 nm and a reference wavelength of 630 nm. Cell viability was calculated by the following formula:

% Cell viability = (A_t/A_s) × 100%

A_t and A_s denoted the absorbance of the test substances and solvent control, respectively.

Detection of apoptosis via APC Annexin V/7-AAD staining assay

Apoptotic cells were evaluated by flow cytometry. MCF-7 cells were treated with various concentrations of 5 or 0.1% DMSO for 48 h. For flow cytometry, 1 × 10⁶ MCF-7 cells in 500 μL binding buffer were stained with 5 μL Annexin V-FITC and 5 μL 7-AAD at room temperature for 15 min. Cells were then analyzed by flow cytometry (488 nm excitation and 600 nm emission filters) using BD FACSCalibur flowcytometer (Becton & Dickinson Company, Franklin Lakes, NJ). Cells undergoing apoptosis are both APC Annexin V positive and 7-AAD negative.

Anti-inflammatory activities

All compounds were evaluated for their inhibitory effects on NO production in lipopolysaccharide-activated RAW 264.7 macrophages as described in the literature.⁶ The cells were obtained from the Cell Bank of the Shanghai Institute of Cell Biology. L-NMMA were used as positive control, and the experiments were conducted for three independent replicates.

Acknowledgements

This research work was funded by the National Natural Science Foundation of China (81503218), the Priority Academic Program Development of Jiangsu Higher Education Institutions (PAPD), the Program for Changjiang Scholars and Innovative Research Team in University (IRT_15R63), and the youth fund project of basic research program of Jiangsu Province (Natural Science Foundation, BK20130651).

References

1. G. Hua and R. W. John, Org. Lett., 2006, 8, 2253–2255.
2. P. G. Bradford and A. B. Awad, Biofactors, 2010, 36, 241–247.
3. E. Ryan, J. Chopra, F. McCarthy, A. R. Maguire and N. M. O'Brien, Br. J. Nutr., 2005, 94, 443–451.
4. B. Sudhamalla, M. Gokara, N. Ahalawat, D. G. Amooru and R. Subramanyam, J. Phys. Chem. B, 2010, 114, 9054–9062.
5. E. Hovenkamp, I. Demonty, J. Plat, D. Lutjohann, R. P. Mensink and E. Trautwein, Prog. Lipid Res., 2008, 47, 37–49.
6. Y. Li, M. Ishibashi, M. Satake, X. Chen, Y. Oshima and Y. Ohizumi, J. Nat. Prod., 2003, 66, 696–698.
7. M. H. Yang, J. S. Wang, J. G. Luo, X. B. Wang and L. Y. Kong, Bioorg. Med. Chem., 2011, 19, 1409–1417.
8. M. H. Yang, J. S. Wang, J. G. Luo, X. B. Wang and L. Y. Kong, J. Nat. Prod., 2009, 72, 2014–2018.
9. M. H. Yang, J. S. Wang, J. G. Luo, X. B. Wang and L. Y. Kong, Can. J. Chem., 2012, 90, 199–204.
10. S. R. Jones, B. S. Selinsky, M. N. Rao, X. H. Zhang, W. A. Kinney and F. S. Tham, J. Org. Chem., 1998, 63, 3786–3789.
11. M. Ishibashi, E. Yamagishi and J. Kobayashi, Chem. Pharm. Bull., 1997, 45, 1435–1438.
12. H. D. Flack and G. Bernardinelli, Acta Crystallogr., Sect. A: Found. Crystallogr., 1999, 55, 908–915.
13. Y. Zhang, J. S. Wang, Y. C. Gu, X. B. Wang and L. Y. Kong, Tetrahedron, 2014, 70, 6594–6606.
14. Z. L. Hong, J. Xiong, S. B. Wu, J. J. Zhu, J. L. Hong, Y. Zhao, G. Xia and J. F. Hu, Phytochemistry, 2013, 86, 159–167.
15. K. Machida and M. Kikuchi, Chem. Pharm. Bull., 1999, 47, 692–694.
16. M. Gorecki, E. Jablonska, A. Kruszewska, A. Suszczynska, L. Z. Urbanczyk, M. Gerards, J. W. Morzycki, W. J. Szczepek and J. Frelek, J. Org. Chem., 2007, 72, 2906–2916.
17. G. Snatzke, Angew. Chem., Int. Ed. Engl., 1979, 18, 363–377.
18. B. L. Di, G. Pescitelli, C. Pratelli, D. Pini and P. Salvadori, J. Org. Chem., 2001, 66, 4819–4825.
19. H. Tsutomu, Y. Yukie, A. Takamasa and S. Yumi, Org. Biomol. Chem., 2007, 5, 792–801.
20. P. Khiev, O. K. Kwon, H. H. Song, S. R. Oh, K. S. Ahn, H. K. Lee and Y. W. Chin, Chem. Pharm. Bull., 2012, 60, 955–961.
21. S. Weber, J. Puripattanavong, V. Brecht and A. W. Frahm, J. Nat. Prod., 2000, 63, 636–642.
22. I. S. Lee, S. R. Oh, K. S. Ahn and H. K. Lee, Chem. Pharm. Bull., 2001, 49, 1024–1026.
23. X. H. Cai, X. D. Luo, J. Zhou and X. J. Hao, Org. Lett., 2005, 7, 2877–2879.
24. F. Zhang, J. S. Wang, Y. C. Gu and L. Y. Kong, J. Nat. Prod., 2010, 73, 2042–2046.
25. T. Bifu and R. Stevenson, J. Chem. Soc., Perkin Trans. 1, 1978, 7, 360–363.
26. H. Chavez, A. E. Braun, A. G. Ravelo and A. G. Gonzalez, Tetrahedron, 1997, 53, 6465–6472.
27. O. Kenny, Y. O'Callaghan, N. M. O'Connell, F. O. McCarthy, A. R. Maguire and N. M. O'Brien, J. Agric. Food Chem., 2012, 60, 5952–5961.
28. Y. O'Callaghan, F. O. McCarthy and N. M. O'Brien, Biochem. Biophys. Res. Commun., 2014, 446, 786–791.
29. M. Tasyriq, I. A. Najmuldeen, L. L. In, K. Mohamad, K. Awang and N. Hasima, J. Evidence-Based Complementary Altern. Med., 2012, 2012, 765316–765342.
30. H. Li, Y. Li, X. B. Wang, T. Pang, L. Y. Zhang, J. Luo and L. Y. Kong, RSC Adv., 2015, 5, 40465–40474.
31. G. M. Sheldrick, SHELXS-97, Program for Crystal Structure Resolution, University of Göttingen, Germany, 1997.
32. G. M. Sheldrick, SHELXL-97, Program for Crystal Structure Refinement, University of Göttingen, Germany, 1997.
33. Z. B. Zhou, J. G. Luo, K. Pan, S. M. Shan, W. Zhang and Y. K. Ling, Planta Med., 2013, 79, 1730–1735.

---

Footnote

† Electronic supplementary information (ESI) available. CCDC 1433008. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c5ra23626f

---