Novel cycloartane triterpenoid from Cimicifuga foetida (Sheng ma) induces mitochondrial apoptosis via inhibiting Raf/MEK/ERK pathway and Akt phosphorylation in human breast carcinoma MCF-7 cells

Rhizome of C. foetida was collected in Emei Mountain, Sichuan province, China in October 2012 and identified by Dr. Sibao Chen by comparing its morphological features [8]. A voucher specimen (SZRI20121045) was deposited in the herbarium of state key laboratory of Chinese medicine and molecular pharmacology.

Infrared spectra (IR) were recorded on a Shimadzu IR-450 spectrometer (Shimadzu, Kyoto, Japan). High resolution fast atom bombardment mass spectra (HR-FAB-MS) were recorded on a VG-Autospec-3000 spectrometer (Micromass, Manchester, UK), in m/z (percentage relative intensity of base peak). The nuclear magnetic resonance (NMR) spectra were measured in pyridine-d ₅ on a Bruker AM-400 spectrometer (BrukerDaltonics, MA, USA), using tetramethyl silane as internal standard. Silica gel-_60H (100–200 and 200–300 mesh) and silica gel GF₂₅₄ sheets (0.20–0.25 mm; both from Qingdao Haiyang Chemical Group Co., Qingdao, China) were used for column chromatography (CCG) and thin layer chromatography (TLC), respectively.

3-(4, 5)-Dimethylthiazol-2-yl)-2, 5-diphenyl tetrazolium bromide (MTT), Hoechst 33258, propidium iodide (PI), doxorubicin (Dox) and carbonyl cyanide 3-c hlorophenylhydrazone were purchased from Sigma-Aldrich (St. Louis, MO, USA). 5, 5′, 6, 6′-tetrachloro-1, 1′, 3, 3′-tetraethyl benzimidazolyl-carbocyanine iodide (JC-1) and Alexa Fluor^® 488 Annexin V/dead cell apoptosis kit were obtained from invitrogen (Carlsbad, CA, USA). BCA protein assay kit and electro chemiluminescence were obtained from Thermo Fisher Scientific (Rockford, IL, USA). Antibodies against cyclin B1, CDK1, p-CDK1(Thr¹⁶¹), caspase-9, cleaved caspase-9, PARP, cleaved PARP, Akt, p-Akt (Thr³⁰⁸ and Ser⁴⁷³), p-ERK1/2 and p-c-Raf were from cell signaling technology (Beverly, MA, USA). Antibodies against Bcl-2 and Bax were bought from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Protease inhibitor cocktail tablets and phosphatase inhibitor cocktail tablets were supplied by Roche (Roche Applied Science, Mannheim, Germany). Dimethylsulfoxide (DMSO); RPMI-1640 and Dulbecco’s modification of Eagle’s medium (DMEM), phosphate-buffered saline (PBS), fetal bovine serum (FBS), penicillin-streptomycin (P/S) and trypsin-ethylene diamine tetraacetic acid (EDTA) were purchased from life technologies (Grand Island, NY, USA).

The rhizome of C. foetida (0.5 kg) was extracted three times with 95 % EtOH for 1 h under reflux. After combination of extractions and solvent removal, the residue (129 g) was suspended in water (1 L) and partitioned successively with 200 mL each of petroleum ether (60–80 ℃), ethyl acetate and n-BuOH. The ethyl acetate fraction (28 g) was subjected to CCG on silica gel-_60H (100–200 mesh). Gradient elution with CHCl₃–MeOH (1:0, 50:1, 20:1, 10:1 and 0:1), obtained five fractions: A (3.7 g), B (4.9 g), C (6.28 g), D (1.5 g) and E (1.8 g).

Fraction D was subjected to CCG on silica gel-_60H (200–300 mesh), eluted with CHCl₃–MeOH (70:30) and further purified by Sephadex G₁₀ CCG, eluted with MeOH to obtain ADHC-AXpn (15 mg).

Fraction C was subjected to repeated CCG on silica gel-_60H (200–300 mesh), eluted with CHCl₃ and acetone (3:1) to give sub-fractions 1–5. Sub-fraction 5 was subjected to CCG on silica gel-_60H (200–300 mesh), eluted in gradient with CHCl₃-MeOH (90:10, 85:15 and 80:20). The fraction eluted by CHCl₃-MeOH (80:20) was purified by Sephadex G₁₀ CCG, and eluted with MeOH, to give DHC-Xpn (40 mg). Chemical structures of ADHC-AXpn and DHC-Xpn were elucidated by their spectral data (IR, MS and ¹H, ¹³C-NMR) and by comparison with data in the literature. The purity of isolated components was determined over 98 % by peak area normalization method in HPLC analysis by an Agilent 1200 liquid chromatography system (HP Agilent Technologies, Palo Alto, CA, USA).

Cell lines MCF-7 (estrogen receptor-positive phenotype), HepG2, HeLa and PC3 cells were obtained from American type culture collection (Manassas, VA, USA). HepG2/ADM cells (multidrug resistance phenotype) were kindly provided by Prof. Kwok-Pui Fung (The Chinese University of Hong Kong, Hong Kong, China). Human MCF10A mammary epithelial cells were obtained from Invitrogen (Carlsbad, CA, USA). The MCF-7, HepG2, HepG2/ADM and PC3 cells were cultured in RPMI-1640 medium supplemented with 10 FBS and 1 % (v/v) P/S at 37 ℃ in a humidified incubator containing 5 % CO₂. HeLa cells were cultured in DMEM medium with the same culture condition mentioned above. Human MCF10A mammary epithelial cells were cultured in a condition mentioned previously [11]. HepG2/ADM cells were cultured with 1.2 μM of Dox; during cell passage to keep their multidrug resistance property as compared with the corresponding parental cells.

The inhibitory effects of ADHC-AXpn and DHC-Xpn on the growth of tested cells were evaluated by MTT assay. Briefly, all tested cells (0.8 × 10⁴/well) were seeded in 96-well plates, cultured for 24 h, then exposed to different concentrations of ADHC-AXpn and DHC-Xpn, respectively for another 24 or 48 h. Subsequently, 30 μL of 5 mg/mL MTT dissolved in PBS was added to each well and incubated for 4 h after removing the media. Formazan crystals were dissolved with 100 μL of DMSO, with absorbance at 490 nm measured by Bio-Rad 680 microplate reader (Bio-Rad, CA, USA). Cell viability was determined as a percentage of the control, with Dox as positive control.

MCF-7 cells (2.5 × 10⁵/well) were seeded in 6-well plates, cultured for 24 h, then exposed to different concentrations of ADHC-AXpn (20, 30 or 35 μM) for 48 h before digestion in 0.25 % trypsin to reconstitute single-cell suspensions. Cell suspensions were then transferred into 6-well plates (600 cells/well) and incubated at 37 ℃ for 12 days. The formed colonies were fixed in 4 % paraformaldehyde for 30 min and stained with 1 % crystal violet for 5 min. These experiments were performed three times. Colonies with diameters larger than 0.5 mm were counted to calculate the colony formation rate.

The nuclear morphology of cells treated with ADHC-AXpn was observed with a Hoechst 33258 staining assay. MCF-7 cells (2.5 × 10⁵/well) were seeded in 6-well plates, cultured for 24 h, and treated with different concentrations of ADHC-AXpn (20, 30 or 35 μM) for 48 h. The cells were then washed three times with PBS, fixed in 4 % paraformaldehyde at 4 ℃ for 30 min, and stained with 10 μg/mL of Hoechst 33258 for 15 min at 37 ℃. Changes in nuclear morphology were monitored with an Olympus IX51 inverted microscope (Olympus Corporation of the Americas, Inc., Central Valley, PA, USA).

Cell cycle distribution was analyzed by a fluorescent probe propidium iodide (PI) staining assay. MCF-7 cells (2.5 × 10⁵/well) were seeded in 6-well plates, cultured for 24 h, then treated with different concentrations of ADHC-AXpn (20, 30 or 35 μM) for 24 or 48 h. Cells were then collected and fixed with 75 % ethanol at 4 ℃ overnight. After centrifugation in an Eppendorf 5417R centrifuge (Eppendorf Corp., Hamburg, Germany) at 800g for 5 min, the cells were incubated with PBS containing 0.02 mg/mL PI and 0.1 mg/mL RNase A in darkness at 37 ℃ for 30 min. The emitting fluorescence was measured by Guava Easy Cytometer (Guava Technologies, Millipore, Billerica, MA, USA); data were analyzed with MultiCycle AV software (Phoenix Flow Systems, San Diego, CA, USA).

MCF-7 cells (2.5 × 10⁵/well) were seeded in 6-well plates, cultured for 24 h, and exposed to differing concentrations of ADHC-AXpn (20, 30 or 35 μM) for 36 h. Cells were then collected and stained by Alexa Fluor^® 488 Annexin V/dead cell apoptosis kit according to manufacturer instructions. Green fluorescence emitted from Annexin V-FITC and red fluorescence from PI were detected by Guava Easy Cytometer (Guava Technologies, Millipore, Billerica, MA, USA). Data were analyzed by FlowJo 7.6 software (TreeStar, SanCarlos, CA, USA).

Changes in MMP were measured using a lipophilic cationic fluorescent probe JC-1. Briefly, MCF-7 cells (2.5 × 10⁵/well) were seeded in 6-well plates, cultured for 24 h, then treated with ADHC-AXpn (20, 30 or 35 μM) for 24 h. The cells were collected and incubated with 5 μM JC-1 in darkness at 37 ℃ for 15 min, then assayed with a Guava Easy Cytometer (EasyCyte 8 HT, Millipore, USA) to evaluate changes in MMP..

MCF-7 cells (2.5 × 10⁵/well) were seeded in 100-mm culture dishes, cultured for 24 h, then incubated with differing concentrations of ADHC-AXpn for various times. Cells were then collected by trypsinization with 0.25 % Trypsin-EDTA and transferred to centrifuge tubes. After centrifugation at 800g for 5 min, cells were washed twice with ice-cold PBS and then lysed in RIPA buffer containing 0.5 M DTT, 0.1 M PMSF and 20× phosphatase inhibitor for 30 min on ice. Protein samples were harvested by centrifugation at 12,000g at 4 ℃ for 12 min. Protein contents were quantified with BCA Protein Assay Kit (Thermo Scientific Pierce, Rockford, USA). The protein samples (35 μg) along with a rainbow-colored protein molecular marker (Amersham, Buckinghamshire, UK) were separated by 10 % SDS-PAGE and transferred to polyvinylidene fluoride (PVDF) membranes (Millipore, Bedford, MA, USA) by a semi-dry transfer apparatus (Bio-Rad, CA, USA). The membranes were blocked with 5 % nonfat dry milk in Tris-buffered saline containing 0.1 % Tween-20 (TBS-T) at room temperature for 1 h and incubated overnight with primary antibodies containing 5 % BSA at 4 ℃. Membranes were washed with TBS-T three times and appropriate secondary antibodies conjugated with horseradish peroxidase were added and incubated for another hour. Finally, immunoblots were visualized by enhanced chemiluminescence detection reagents, with β-actin as a loading control.

ADHC-AXpn was isolated as a white powder. The HR-FAB-MS showed a pseudo-molecular ion peak at m/z 725.3887 [M + Na]⁺ (calc. 725.3893), which supported the molecular formula as being C₃₉H₅₈O₁₁. The IR spectrum displayed the presence of hydroxyl groups (3338 cm⁻¹) and a carbonyl group (1730 cm⁻¹). The ¹H-NMR spectrum (Table 1) showed signals of cyclopropane methylene groups at δ 0.43 and 0.96 ppm (each ¹H, d, J = 4.0 Hz), six methyl groups at δ 1.05, 1.08, 1.13, 1.20, 1.65 and 1.67 ppm; an aromatic proton at δ 4.82 ppm (d, J = 8.0 Hz), and a series of overlapped signals, which suggested that ADHC-AXpn was a cycloartane-type triterpene glycoside [12]. The ¹³C-NMR (Table 1) displayed the diagnostic signals of two oxygen-bearing alkyl carbons at δ 87.0 (C-24) and 72.1 ppm (C-23), and an oxygen-bearing quaternary carbon at δ 113.0 ppm (C-16), which implied that ADHC-AXpn was a cimigenol-type triterpene [13]. By fully comparing ¹H and ¹³C NMR spectral data of ADHC-AXpn with those of the known component 25-O-acetyl-7,8-didehydrocimigenol-3-O-β-d-xylopyranoside (25-AC) [12], the structure of ADHC-AXpn was similar to that of 25-AC, excepting an additional acetyl moiety. When the ¹³C-NMR spectrum of ADHC-AXpn was compared with that of 25-AC, the signals of C-1′ and C-3′ were seen to shift to higher field by 2.6 and 2.0 ppm, while the signal of C-2′ shifted to lower field. These chemical shift variations suggested an acetyl moiety was linked to the C-2′ of the xylose unit [12]. Locations of the two acetyl groups were directly confirmed by hetero-nuclear multiple bond correlation (HMBC) experiments (Fig. 2). Briefly, significant correlation was observed between H-2′ (δH 5.56 ppm, dd, J = 7.5, 8.1 Hz) and the carbonyl signal at δ170.3 ppm, which indicated an acetyl group was assignable to C-2′. The methyl signals at δH 1.65 ppm (Me-26)/δH 1.67 ppm (Me-27) also showed correlations with a quaternary carbon signal at δC 83.44 ppm (C-25), the methine carbon signal at δC 87.05 ppm (C-24), and the methyl carbon signals at δC 21.58 ppm (C-27)/δC 24.14 ppm (C-26), which indicated that another acetyl group was located at C-25. A complete ¹H and ¹³C-NMR spectral assignment of ADHC-AXpn was summarized in Tables 1 and 2.

Taken together, the structure of ADHC-AXpn was formulated as 25-O-acetyl-7,8-didehydrocimigenol-3-O-β-d-(2-acetyl) xylopyranoside (Fig. 1). The known DHC-Xpn was determined as 7,8-didehydrocimigenol-3-O-β-d-xylopyranoside by comparison of its spectral data with those reported in the literature [14].

The inhibitory effects of ADHC-AXpn and DHC-Xpn on the viability of various carcinoma cells were determined by MTT assay. ADHC-AXpn was much more potent than DHC-Xpn in decreasing the viability of the tested cells (MCF-7, HepG2/ADM, HepG2, HELA and PC3 cells) in both 24 and 48 h treatment periods apart from PC3 cells (Table 2). Among these cancer cells treated with ADHC-AXpn, MCF-7 cells appeared to be more sensitive to ADHC-AXpn than other cancer cells, as indicated by the IC₅₀ values (Fig. 3a). As shown in Fig. 3b, the colony formation assay also confirmed that ADHC-AXpn (20, 30 or 35 μM) could inhibit proliferation of MCF-7 cells (P = 0.0078, P < 0.001, P < 0.0001). The IC₅₀ values of ADHC-AXpn in human MCF10A mammary epithelial cells at 48 h were 78.63 ± 3.26 μM, which was much higher than for cancer cell line MCF-7. These data indicated that ADHC-AXpn was selectively cytotoxic for cancer cells.

We investigated the effect of ADHC-AXpn on the MCF-7 cell cycle by flow cytometry. Treating MCF-7 cells with ADHC-AXpn for 24 h induced a G₂/M arrest (Fig. 4a). The percentage of cells in subG₁ phase increased when ADHC-AXpn exposure was extended to 48 h. To explore the mechanism of ADHC-AXpn-induced G₂/M arrest, we investigated its effects on G₂/M regulatory proteins. Western blotting showed ADHC-AXpn caused a time- and dose-dependent decrease in cyclin B1 (Fig. 4b). ADHC-AXpn decreased cyclin-dependent kinase 1 (CDK1), which regulates the cell cycle by forming the CDK-cyclin complex, and phospho-CDK1 (Thr¹⁶¹). ADHC-AXpn arrests the cell cycle in G₂/M phase by decreasing expression of cyclin B1, CDK1 and phospho-CDK1 (Thr¹⁶¹).

Changes of cellular morphology were examined by fluorescent microscopy analysis (Fig. 5a). After treatment with different concentrations of ADHC-AXpn (20, 30 or 35 μM) for 48 h, cells stained with Hoechst 33258 showed apoptotic features, such as cytoplasmic shrinkage and condensation of nuclear chromatin. Consistently, PI/Annexin V double staining analysis (Fig. 5b) showed that treatment with ADHC-AXpn (30 or 35 μM) for 48 h led to significantly increased cell percentages in early and late apoptotic phases (P = 0.0094 and P = 0.0042 vs. control, respectively). Cleavage of PARP was also seen in ADHC-AXpn-treated MCF-7 cells (Fig. 5c).

Changes of MMP were measured by a fluorescent dye (JC-1) to further investigate the apoptotic pathway. MCF-7 cells treated with ADHC-AXpn (20, 30 or 35 μM) for 24 h showed a gradual decrease in the percentage of cells with high MMP from 94.1 (control) to 68.7 %, 63.0 and 55.9 %, respectively (Fig. 6a). We examined expression levels of pro-apoptotic protein Bax and anti-apoptotic protein Bcl-2. Western blot analysis showed that ADHC-AXpn treatment down-regulated Bcl-2 and clearly up-regulated Bax (Fig. 6b). The precursor form of caspase-9 was decreased, whereas expression of cleaved caspase-9 was increased. ADHC-AXpn exposure remarkably decreased p-Akt (Thr³⁰⁸ and Ser⁴⁷³), although total Akt showed no visible change.

The Raf/MEK/ERK pathway is beneficial for survival of MCF-7 cells, and blockage of this pathway can trigger apoptosis [15]. To investigate whether the Raf/MEK/ERK pathway affects ADHC-AXpn-induced apoptosis, we assessed expression levels of p-ERK1/2 and p-c-Raf. Compared with the control group, ADHC-AXpn treatment decreased phosphorylation of ERK1/2 in both time- and dose-dependent manners (Fig. 6c). The upstream regulator of ERK1/2, p-c-Raf, was also down-regulated after exposure to ADHC-AXpn.