On the basis of cytotoxicity assay results of different extracts and pure compounds against human normal cell lines L-O2 and GES-1 (Table 4), 12 terpenoids were isolated from the EtOAc extract of kansui. By spectral and physiochemical data analysis and/or comparison with literatures data, their structures were elucidated as kansuinines A, B, C (1–3) [17,21], kansuiphorin C (4) [20], 3-O-(2′E,4′Zdecadienoyl)-20-O-acetylingenol (5) [17], 3-O-(2′E,4′E-decadienoyl)-20-O-acetylingenol (6) [17], 3-O-(2′E,4′Z-decadienoyl)-20-deoxyingenol (7) [17,23], 3-O-benzoyl-20-deoxyingenol (8) [22], 5-O-benzoyl-20-deoxyingenol (9) [17], kansenone (10) [17], epi-kansenone (11) [17], and euphol (12) [17,21] (Figure 1). This was the first report of the bio-guided isolation of different fractions and pure terpenoids from kansui extracts by the gastrointestinal and hepatic cytotoxicity assay against human normal cell lines L-O2 and GES-1.

Kansuinine A (1): white powder, mp 218–220 °C; R_f = 0.25 (SiO₂, petroleum ether-chloroform-EtOAc = 5:3:3); UV (MeOH) λ_max 232 nm; ESI-MS (positive) m/z 753 [M + Na]^{+; 1}H-NMR data, see Table 1.

Kansuinine B (2): colorless crystal; mp 215–216 °C; R_f = 0.30 (SiO₂, petroleum ether-chloroform-EtOAc = 5:3:3); UV (MeOH) λ_max 232 nm; ESI-MS (positive) m/z 745 [M + Na]^{+; 1}H-NMR data, see Table 1.

Kansuinine C (3): colorless needle crystal; mp 217–218 °C; R_f = 0.28 (SiO₂, petroleum ether-chloroform-EtOAc = 5:3:3); UV (MeOH) λ_max 232 nm; ESI-MS (positive) m/z 745 [M + Na]^{+; 1}H-NMR data, see Table 1.

Kansuiphorin C (4): colorless gum; R_f = 0.83 (SiO₂, petroleum ether-chloroform-EtOAc = 5:3:1); UV (MeOH) λ_max 230; ESI-MS (positive) m/z 501 [M + Na]^{+; 1}H-NMR data, see Table 2.

3-O-(2′E,4′Z-Decadienoyl)-20-O-acetylingenol (5): colorless gum; R_f = 0.55 (SiO₂, petroleum ether-chloroform-EtOAc = 5:3:2); UV (MeOH) λ_max 266, 206; ESI-MS (positive) m/z 563 [M + Na]^{+; 1}H-NMR data, see Table 2.

3-O-(2′E,4′E-Decadienoyl)-20-O-acetylingenol (6): colorless gum; R_f = 0.65 (SiO₂, petroleum ether-chloroform-EtOAc = 5:3:2); UV (MeOH) λ_max 266, 203; ESI-MS (positive) m/z 563 [M + Na]^{+; 1}H-NMR data, see Table 2.

3-O-(2′E,4′Z-Decadienoyl)-20-deoxyingenol (7): colorless gum; R_f = 0.66 (SiO₂, petroleum ether-chloroform-EtOAc = 5:3:2); UV (MeOH) λ_max 260, 206; ESI-MS (positive) m/z 505 [M + Na]^{+; 1}H-NMR data, see Table 2.

3-O-Benzoyl-20-deoxyingenol (8): colorless gum; R_f = 0.72 (SiO₂, petroleum ether-chloroform-EtOAc = 5:3:2); UV (MeOH) λ_max 273, 229, 208; ESI-MS (positive) m/z 459 [M + Na]^{+; 1}H-NMR data, see Table 2.

5-O-Benzoyl-20-deoxyingenol (9): colorless gum; R_f = 0.61 (SiO₂, petroleum ether-chloroform-EtOAc = 5:3:2); UV (MeOH) λ_max 273, 229, 208; ESI-MS (positive) m/z 459 [M + Na]^{+; 1}H-NMR data, see Table 2.

Kansenone (10): colorless gum; R_f = 0.61 (SiO₂, petroleum ether-chloroform-EtOAc = 5:3:2); UV (MeOH) λ_max 252; ESI-MS (positive) m/z 441 [M+H]^{+; 1}H-NMR data, see Table 3.

Epi-kansenone (11): colorless gum; R_f = 0.64 (SiO₂, petroleum ether-chloroform-EtOAc = 5:3:2); UV (MeOH) λ_max 252; ESI-MS (positive) m/z 441 [M + H]^{+; 1}H-NMR data, see Table 3.

Euphol (12): white crystal (ethyl acetate); mp 116–118 °C; R_f = 0.52 (SiO₂, hexane-acetone-glacial acetic acid = 8:1:0.1); UV (MeOH) λ_max 200; ¹H-NMR (300 MHz, CDCl₃) δ 0.75 (3H, s, 30-H), 0.80 (3H, s, 18-H), 0.84 (3H, d J = 3.3 Hz, 21-H), 0.87 (3H, s, 28-H), 0.95 (3H, s, 19-H), 1.00 (3H, s, 29-H), 1.60 (3H, s, 27-H), 1.68 (3H, s, 26-H), 3.24 (1H, dd J = 4.8, 11.6 Hz, 3α-H), 5.09 (1H, t, J = 7.5 Hz, 24-H); ¹³C-NMR (75 MHz, CDCl₃): 35.4 (C-1), 27.7 (C-2), 78.9 (C-3), 37.3 (C-4), 50.9 (C-5), 18.9 (C-6), 27.9 (C-7), 134.0 (C-8), 133.5 (C-9), 38.9 (C-10), 21.5 (C-11), 28.1 (C-12), 44.1 (C-13), 49.9 (C-14), 30.9 (C-15), 29.8 (C-16), 49.6 (C-17), 15.6 (C-18), 20.1 (C-19), 35.8 (C-20), 18.9 (C-21), 35.2 (C-22), 24.7 (C-23), 125.2 (C-24), 130.8 (C-25), 25.7 (C-26), 17.7 (C-27), 24.4 (C-28), 28.0 (C-29), 15.5 (C-30).

As shown in Table 4, the EtOH and EtOAc extract exhibited significant inhibition of cell proliferation against two human normal cell lines L-O2 and GES-1 with dose-dependent relationship. The IC₅₀ values of EtOH extract on L-O2 and GES-1 were 42.02 and 30.67 μg/mL, respectively. And the cytotoxicity of EtOAc extract on L-O2 and GES-1 were stronger than that of EtOH extract. The IC₅₀ values of EtOAc extract on L-O2 and GES-1 were 27.08 and 21.89 μg/mL, respectively. The water extract showed no obvious cytotoxicity against these two human normal cell lines.

Compared with the control group, compounds 5, 7, 8 and 10 in the concentration range of 0.78–12.5 μg/mL, compounds 6 and 11 in the concentration range of 0.39–6.25 μg/mL, and compound 9 in the concentration range of 1.56–25 μg/mL, showed significant inhibition activity on GES-1 cell lines growth with a dose-dependent relationship (Figure 2). And compounds 5, 6, 7, 10 and 11 in the concentration range of 0.78–12.5 μg/mL, compound 9 in the concentration range of 1.56–25 μg/mL, and compound 8 in the concentration range of 3.125–50 μg/mL, had significant inhibition activity on L-O2 cell lines growth with a dose-dependent relationship (Figure 3). Whereas, other compounds did not show the obvious cytotoxicity activity in the concentration range of 3.125–50 μg/mL. Considering their structure, the results indicated that ingenane-type diterpenoids (5–9) except compound 4 and 8-ene-7-one triterpenoids (10 and 11) in kansui possessed stronger gastrointestinal toxicity and hepatotoxicity than other compounds. More interestingly, the IC₅₀ values of 5–11 and the kansui extracts on GES-1 cell lines were less than those on L-O2 cell lines, which suggested that the gastrointestinal toxicity of kansui is stronger than its hepatotoxicity. This result coincided with the clinical toxic performance of kansui [14–16].

Analytical grade ethanol, EtOAc and petroleum ether (60–90 °C) (Sinopharm Chemical Reagent Co., Ltd. Shanghai, China) were used for extraction and isolation. The silica gel (Qingdao Ocean chemical industry, Qingdao, China) was used for column chromatography. Dulbecco’s Modified Eagle’s Medium (DMEM) was purchased from Gibco Co., Ltd. (Grand Island, NY, USA); fetal bovine serum (FBS) and calf serum (CS) were purchased from Sijiqing Biological Engineering Material Co., Ltd. (Hangzhou, China); 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) was obtained from Beijing Solarbio science and technology Co., Ltd. (Beijing, China); PBS buffer was purchased from Boster Bio-engineering Co., Ltd. (Wuhan, China); dimethylsulfoxide (DMSO) was purchased from Sinopharm Chemical Reagent Co., Ltd. (A·R grade, Shanghai, China).

Melting points were determined on a Yanagimoto MP-500 D micro-melting point apparatus and are uncorrected. The UV spectrum was obtained in MeOH on a Shimadzu UV-2401 spectrophotometer. The NMR spectra were recorded on Bruker Avance AV 300 instrument operating at 300 MHz for ¹H and 75 MHz for ¹³C, using standard pulse sequences. Chemical shifts are reported on the δ scale in parts per million, with TMS as an internal standard. The electrospray ionization mass spectra (ESI-MS) were obtained on a Q-TOF Micro mass spectrometer (Waters Company, Milford, MA, USA). Thin-layer chromatography (TLC) was performed on Qingdao Ocean TLC plates (0.25 mm thickness, Qingdao Ocean chemical industry, Qingdao, China), with compounds visualized by spraying with 10% (v/v) H₂SO₄ in ethanol solution and then heating on a hot plate. Semipreparative HPLC was performed on a Hanbang NP 7000 apparatus with a NU 3000 detector. A Hanbang Phecda Si silica gel column (20 × 250 mm i.d., 5 μm) was used for preparative purposes. Laminar flow clean bench (Model: SW-CJ-1F) was purchased from Suzhou Purification Equipment Co., Ltd. (Suzhou, China); CO₂ incubator (Model: MCO-20AIC) was purchased from Sanyo Denki Co., Ltd. (Moriguchi, Osaka, Japan); Microplate Reader (model: 680) obtained from Bio-Rad Laboratories (Hercules, CA, USA). Inverted phase contrast microscope (model: CKX31) was purchased from Olympus Co., Ltd. (Tokyo, Japan).

The roots of Euphorbia kansui T.N. Liou ex T.P. Wang were collected from Red River valley of Baoji (Shanxi Province, China). The crude plant was identified by Professor Chungen Wang (Nanjing University of Chinese Medicine, Nanjing, China). The voucher specimens (No. NJUTCM-20091008) was deposited in the Herbarium of Nanjing University of Chinese Medicine, Jiangsu, China.

The dried and crushed roots of E. kansui (20.0 kg) were extracted with 95% EtOH under 50 °C water bath, the supernatant was separated every day, and the solvent was removed under reduced pressure, and the residue (2.35 kg) was partitioned with ethyl acetate and water to provide the EtOAc fraction (1.032 kg) and water fraction (1.056 kg). The EtOAc fraction 0.904 kg was subjected to silica gel column chromatography (100 × 10 cm, 200–300 mesh, eluted with petroleum ether and ethyl acetate in increasing polarity). The column chromatographic eluents (500 mL each) were concentrated under reduced pressure and combined according to TLC monitoring into 10 fractions. Fraction 1 was subjected to silica gel column chromatography eluted with petroleum ether-ethyl acetate 100:1, and was further recrystallized to give compound 12 (a large amount). Fractions 2–4 was subjected to silica gel column chromatography eluted with petroleum ether-ethyl acetate from 100:16 to 100:18, and were further recrystallized to give 1 (132 mg), 2 (86 mg) and 3 (46 mg). Fraction 5 was subjected to silica gel column chromatography eluted with petroleum ether-ethyl acetate 100:2, and was further purified by preparative TLC using silica gel (petroleum ether-chloroform-ethyl acetate 5:3:1, R_f = 0.83) to give compound 4 (106 mg). Fraction 6 was subjected to silica gel column chromatography eluted with petroleum ether-ethyl acetate 100:5, and was further purified by preparative HPLC (petroleum ether-ethyl acetate, ρ 0.705, UV detector set at 260 nm, flow rate at 15 mL/min) to give compound 9 (41 mg, t_R 31.0 min). Fraction 7 was subjected to silica gel column chromatography eluted with petroleum ether-ethyl acetate 100:5, and was further purified by preparative HPLC (petroleum ether-ethyl acetate, ρ 0.700, UV detector set at 260 nm, flow rate at 15 mL/min) to give compound 7 (56 mg, t_R 14.3 min). Fraction 8 was subjected to silica gel column chromatography eluted with petroleum ether-ethyl acetate 100:6, and was further purified by preparative HPLC (petroleum ether-ethyl acetate, ρ 0.692, UV detector set at 260 nm, flow rate at 15 mL/min) to give compound 8 (14 mg, t_R 16.4 min). Fraction 9 was subjected to silica gel column chromatography eluted with petroleum ether-ethyl acetate 100:8, and was further purified by preparative HPLC (petroleum ether-ethyl acetate, ρ 0.706, UV detector set at 260 nm, flow rate at 15 mL/min) to give compound 10 (73 mg, t_R 20.6 min) and 11 (32 mg, t_R 22.7 min). Fraction 10 was subjected to silica gel column chromatography eluted with petroleum ether-ethyl acetate 100:11, and was further purified by preparative HPLC (petroleum ether-ethyl acetate, ρ 0.705, UV detector set at 260 nm, flow rate at 15 mL/min) to give compound 5 (29 mg, t_R 19.0 min) and 6 (21 mg, t_R 20.3 min).

Human normal liver cell L-O2 was kindly donated by the School of Life Science, Nanjing Normal University (Nanjing, China), and Human normal gastric epithelial cell GES-1 was kindly donated by the Department of Gastroenterology, Nanjing Drum Tower Hospital (Nanjing, China). Human normal liver cell LO-2 was routinely cultured in DMEM medium [16], supplemented with 10% (v/v) fetal bovine serum (FBS), 100 U/mL of penicillin and 100 μg/mL of streptomycin in a humidified atmosphere with 5% CO₂ at 37 °C in an incubator. The culture medium was changed every day. Cultures were dissociated with 0.25% trypsase in phosphate buffered (pH 7.2~7.4) saline (PBS). Human normal gastric epithelial cell GES-1 was routinely cultured in DMEM medium [30], supplemented with 15% (v/v) calf serum (CS), 100 U/mL of penicillin and 100 μg/mL of streptomycin in a humidified atmosphere with 5% CO₂ at 37 °C in an incubator. The culture medium was changed every two days. Cultures were dissociated with 0.25% trypsase and 0.05% EDTA in PBS (pH 7.2~7.4).

Cytotoxicity activity was assayed by MTT method as described [31]. Briefly, cells were seeded in 96-well plates (Costar, Corning Co., Ltd., Beijing, China) with 100 μL at a concentration of cell suspension of 1.0 × 10₄ cells/mL (L-O2 cell) and 1.0 × 10₅ cells/mL (GES-1 cell), and incubated for 24 h at 37 °C in a humidified atmosphere of 5% CO₂ in an incubator.

Every sample was solubilized in dimethyl sulfoxide (DMSO) to a concentration of 10 mg/mL and stored at 4 °C, then diluted with complete medium to the desired series concentration before treating the cells. Each solution was added into 96-well plates with 100 μL per well, repeated 6 well. Cells of the control groups were treated with the same volume of medium. The treated cells with different groups were incubated for 48 h at 37 °C in a humidified atmosphere of 5% CO₂ in an incubator. Then, the treated cells were added with 20 μL/well of MTT (5.0 mg/mL), and incubated for a further 4 h in an incubator. The growth medium was removed from all the wells, and 150 μL DMSO were added to each well. The plates were shaken gently for 10 min to blend the mixture. The absorbance value of each well was read at 490 nm using a microplate reader. All experiments were performed at least three times. The inhibition rate of cell proliferation was calculated according to the formula:

The results were expressed as an inhibition ratio and IC₅₀ (concentration that inhibits 50% of cell growth). The data were analyzed with software SPSS 15.0 (SPSS Inc.: Chicago, IL, USA). The anti-proliferation activity is more significant when the IC₅₀ is smaller.