UV and IR spectra were determined on a Shimadzu UV-250 spectrometer and a Shimadzu FTIR-8400S spectrometer, respectively. LC-PDA-ESIMS data were recorded on a Waters ACQUITY SQD MS system (Waters, Milford, MA, United States) connected to a Waters 1525 HPLC with a 2998 Photodiode Array Detector (Waters, Milford, MA, United States) and a Waters Sunfire^TMC18 column (5 μm, 4.6 mm × 150 mm). NMR (MeOH-d4 or DMSO-d6) spectra were acquired on an AVANCE III 600 MHz NMR spectrometer equipped with Micro NMR tubes (1.4 mm). The chemical shifts (δ) were reported in ppm, and coupling constants (J) were given in Hz. The ESIMS and HRESIMS data were recorded on a Q-TOF Micro LC-MS-MS mass spectrometer. A Thermo C18 5 μm column (22 mm × 150 mm) was used for semi-preparative HPLC. A Waters 2535 HPLC fitted with a 2998 Photodiode Array Detector and a 2707 Autosampler was used for the semi-preparative separations. Silica gel (300–400 mesh, Yantai Jiangyou Silica Development, Co., Ltd., Yantai, China) were used for column chromatography. Silica gel GF254 precoated glass plates (1.00 mm, Yantai Jiangyou Silica Development, Co., Ltd., Yantai, China) were used for preparative TLC (PTLC).

The roots of Sophora flavescens Aiton (family Leguminosae) were collected from Lingyuan City, Liaoning province, China in September, 2012, and identified by Professor Dingrong Wan of School of Pharmaceutical Sciences, South-Central University for Nationalities (SCUN), Wuhan, China. Avoucher specimen (No. SC0060) was deposited in School of Pharmaceutical Sciences, SCUN, Wuhan, China.

Air-dried roots of Sophora flavescens Aiton (500 g) were triturated and then extracted sequentially by maceration with n-hexane (4 × 2.0 L, 5 h each) at room temperature, followed by ethyl acetate (4 × 2.0 L, 5 h each) and methanol (4 × 2.0 L, 5 h each). The solvents were evaporated at reduced pressure to yield 4.9, 36.8, and 58.7 g of n-hexane, ethyl acetate (SF-EtOAc), and methanol fractions, respectively.

The EtOAc extract (23 g) was subjected to a normal-phase silica gel column chromatography using a gradient solvent system of CH₂Cl₂-MeOH (1:0→50:1→30:1→20:1→10:1→5:1 →1:1→0:1, containing 0.1% formic acid) to give six major fractions Fr. A (1.5 g), Fr. B (3.3 g), Fr. C (4.2 g), Fr. D (3.9 g), Fr. E (3.1 g), and Fr. F (2.4 g). Fr. B was subject to a normal-phase silica gel column (EtOAc-Acetone, 100:1→30:1 →20:1→10:1→5:1→0:1, containing 0.1% formic acid) to afford four subfractions (B1–B4). 0.8 g of the B3 was further purified by PTLC followed by HPLC (H₂O-CNCH₃, 70%:30%→30%:70%, 40 min, containing 0.1% formic acid in both mobile phases) to give the purified compound and it has been elucidated as kurarinone by comparing its ¹H, ¹³ C NMR and MS data (S1–S3 in Supplementary Material) with the reported literature (Kang et al., 2000).

The NSCLC A549, NCI-H1975 and the human bronchial epithelial BEAS-2B cell lines were purchased from the American Type Culture Collection (ATCC; Manassas, VA, United States). The A549, NCI-H1975, and BEAS-2B cell lines were grown in a DMEM medium (Gibco) supplemented 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin in a humidified atmosphere containing 5% CO₂ at 37°C.

Cell viability of A549, NCI-H1975, and BEAS-2B were analyzed using 3-(4, 5-dimethylthiazol-2-yl)-2, 5-diphenyltetrazolium bromide (MTT) assay (Sai et al., 2015). Cisplatin (CDDP) was used as a positive control, since it is the most commonly prescribed cytotoxic agent for the treatment of NSCLC (Kang et al., 2016). Cells were seeded at 1 × 10⁴ cells/well into 96-well plates and were treated with kurarinone at the indicated dose of 200 μg/mL dissolved in DMSO. After 24 h, 100 uL of MTT solution was added to each well and incubated for 4 h. Then, DMSO was used to dissolve the formazan crystals and the absorbance was measured at 492 nm by a Microplate Reader (BIO-RAD). Then kurarinone was chosen for the further research owing to the good cytotoxicities. Subsequently, cells were incubated with kurarinone at 5, 10, 15, 20 μg/mL and CDDP at 25 μg/mL containing 1‰ DMSO for 12, 24, and 48 h, respectively. And then supernate was abandoned and 100 μL of MTT (5 mg/mL) was added to each well and incubated cells for 4 h. The formazan crystals were dissolved in DMSO and the absorbance was measured by a Microplate Reader (BIO-RAD).

The A549 cells (1 × 10⁵/well) were seeded into 6-well plates, treated by kurarinone (0, 5, 10 μg/mL) and CDDP (25 μg/mL) for 24 h. Subsequently, a phase contrast microscope (Leica, Nussloch, Germany) was used to observe the cellular morphological changes.

The A549 cells (1 × 10⁵/well) were seeded into 6-well plates and then treated with kurarinone (0, 5, 10 μg/mL) and CDDP (25 μg/mL) for 24 h. After discarding the supernatant, 1.0 mL of stationary liquid (methanol: acetic acid = 3:1) were covered for about 30 min. Then, Hoechst 33258 solution (5 μg/mL) was added to the wells for 30 min. And A549 cells were observed under a fluorescence microscope (Leica Microsystems, Wetzlar, Germany). Cells were scored apoptotic if the nuclei presented chromatin condensation, marginalization or nuclear beading (Mo et al., 2013).

The A549 cells (1 × 10⁵ to 4 × 10⁵ cells per well) were seeded in 6-well plates, treated by kurarinone (0, 5, 10 μg/mL) and CDDP (25 μg/mL) for 24 h. And then cells were collected, fixed in 70% ethanol at 4°C overnight. Then the cells were washed in PBS and stained by 100 μL RNase A and 400 μL PI. The cell cycle distribution analysis was measured using a flow cytometer (BD Biosciences, United States) (Shen et al., 2014).

Western blotting assay was used to analyze the expressions of apoptotic proteins in A549 cells in vitro. A549 cells were incubated with different doses of kurarinone for 24 h. Cells were lysed in lysis buffer, and incubated on ice for 30 min. After being centrifuged at 12,000 rpm for 15 min, the protein were separated by electrophoresis on 12% SDS-PAGE and transferred to polyvinylidine difluoride (PVDF) membrane (Bio-Rad). Membranes were blocked in 5% skimmed milk and incubated with primary antibodies Bax, Bcl-2, procaspase-9, procaspase-3, procaspase-7, procaspase-12, Grp78, Akt, Phospho-Akt (Ser473), β-actin, and GAPDH. The incubated mixture was washed with TBST. Then horseradish peroxidase (HRP) secondary antibodies were added with the mixture incubating at 37°C for 1 h. The incubated mixture was washed with TBST. HRP electrogenerated chemiluminescence (ECL) was used to develop, the developed films were taken and rinsed with pure water, and the washed films were dried. Scanning was used for recording.

In order to explore the interaction and illustrate binding model for the active assay, molecular docking was carried out using Surflex-Dock of sybyl-x 1.10 program package (Tripos, St. Louis, MO, United States). The structural information from the theoretically modeled complex may help us to clarify the binding mechanism between Akt (PDB: 4GV1, this 3D crystal structure of Akt is replaced serine with aspartic acid in position 473 for simulation of Ser473 phosphorylation) and kurarinone (Addie et al., 2013). The 3D structure of kurarinone was also generated using SYBYL package. The standard bond lengths and bond angles, geometry optimization was carried out with the help of standard Tripos force field with a distance dependent-dielectric function, energy gradient of 0.001 kcal/mol and Gasteiger-Hückel as the electrostatics. The Ligand-Based Surface Calculation was employed to generate the protomol.

Athymic nu/nu mice (BALB/c), 4–6 weeks of age, were purchased from the Beijing HFK Bioscience, Co., Ltd. (SCXK 2009-0015). Nude mice were implanted subcutaneously on the flank of the mice with 5 × 10⁶ A549 cells (0.1 mL/mouse). Once tumor size reached about 100 mm³, the mice were divided into four groups with 10 mice per group: kurarinone was administered i.p. at doses of 20 and 40 mg/kg/day. The CDDP group was administered i.p. at doses of 2.5 mg/kg/day every 2 days. The control group was injected with the same volume of PBS instead. The tumor volumes were calculated using the following formula: tumor volume (mm³) = 0.56 × length (mm) × width² (square mm). After 27 days injection of kurarinone, mice were killed by cervical dislocation, and the subcutaneous tumors were harvested, weighed and fixed in 10% formalin for further examination.

All data were expressed as mean ± SD from three independent experiments. One-way analysis of variance (ANOVA) was used for multiple group comparisons by Tukey’s post hoc test using GraphPad Prism 5.0 software package. P-values < 0.05 were considered significant.

Kurarinone (Figures 1A,B) was primarily tested for the percentages of growth inhibition on NSCLC A549 and NCI-H1975 cell lines using MTT assay. As presented in Figure 1B, kurarinone was potent to inhibit A549 cells proliferation with low toxic effects on the human bronchial epithelial cell line BEAS-2B (IC₅₀ > 50 μg/mL). From the results, it can be deduced that kurarinone suppressed cell viability on A549 cells in time- and dose-dependent manners. In order to confirm this conjecture, the NSCLC cell line A549 was treated with kurarinone (0, 5, 10, 15, 20 μg/mL) and CDDP (25 μg/mL). After 12, 24 and 48 h of incubation, cell viability, and cytotoxicity were tested and shown in Figures 1C–G.

After incubation for 24 h, we evaluate the influence of kurarinone on A549 cells such as distortion, membrane blebbing, and shrinkage by morphological observation with a phase contrast microscope. Results indicated that the shape of a majority of cells progressively showed shrinkage and necrosis (Figure 2A). Then the cells were identified by Hoechst 33258 staining and cells treated with kurarinone showed morphological changes characteristic of apoptosis, including chromatin condensation and nuclear DNA fragmentation (Figure 2A).

To explore the capability of kurarinone in triggering NSCLC cell apoptosis, A549 cells were treated with 5 and 10 μg/mL of kurarinone for 24 h. Then, the ratios of sub-G1 DNA in the cell population were determined by FACS. The results shown that A549 cells treated with kurarinone for 24 h dose-dependently promoted the percentages of sub-G1 DNA (Figure 2B). These data indicate that kurarinone is able to break cell cycle progression and prevent the cells from potentially becoming cancerous.

In the course of apoptosis, cyt C released from mitochondria to cytoplasm to down-regulate Bcl-2 expression and up-regulate Bax expression. During the apoptotic signaling network, caspases also play an important part, and apoptotic pathways relied on activation of caspases for the final execution of apoptosis (Zhao et al., 2016). Hence, we used western blotting analysis to verify the effect of apoptosis-related proteins by kurarinone (S4 in Supplementary Material). We found that kurarinone can up-regulated pro-apoptosis proteins cleaved-caspase-3, cleaved-caspase-9, Bax, and down-regulated antiapoptosis protein Bcl-2 expression (Figures 3A–C). It demonstrated that a mitochondria-dependent pathway is involved in kurarinone-induced apoptosis in A549 cells.

Recent studies have also pointed out that the signaling pathway of apoptosis mediated by ERS was a new mechanism of apoptosis and ectopically expressed GRP78, caspase-7, and caspase-12 can form a complex to couple ER stress to the cell death program. In our study, we tested the protein levels of Grp78, procaspase-7, and procaspase-12 via western blot analysis (S5 in Supplementary Material). The results indicated that kurarinone down-regulated the expression of Grp78, procaspase-7, and procaspase-12 in a dose-dependent manner (Figures 4A,B).

AKT protein as a vital component of PI3K-AKT pathway is another important regulator of cellular proliferation and survival, and its dysregulation is associated with the development of cancer. Phosphorylation on Ser473 is essential for activation of AKT. In our study, we measured its phosphorylation level via western blot (S5–S6 in Supplementary Material). Kurarinone dose-dependently repressed the phosphorylation of AKT on Ser473 but did not affect protein levels of AKT (Figure 4C).

To further elucidate the potential target for kurarinone, the interaction between kurarinone and Akt was analyzed by Surflex-Dock. The results indicated that the parameter set for the Surflex-Dock simulation was reasonable. From the docking results, we found that kurarinone bound with the kinase hinge region and the lavandulyl was very important to fill the P-loop hydrophobic pocket of AKT (Figure 5). It demonstrated that p-AKT might be the potential antitumor target of kurarinone.

The in vivo anticancer effect of kurarinone was assessed in tumor xenografted mouse model. A549 cells were injected on the frank of the mice subcutaneously to reach about the size of 100 mm³, and then the mice were assigned to four groups randomly: control group with PBS, kurarinone treated group (20 and 40 mg/kg/day), and CDDP treated group (2.5 mg/kg, i.p. every 2 days). On the last day of kurarinone treatment, the tumor volumes significantly reduced compared with the control group (Figures 6A,B,D). Compared with control, kurarinone at the concentrations of 20 and 40 mg/kg significantly decreased the mean tumor weight (P < 0.05 and P < 0.01) (Figure 6C). Kurarinone showed no detectable toxicity in all the groups since there were no statistically significant effects on body weight (Figure 6E), behavior, and appearance between the kurarinone treated groups and control group. To investigate whether kurarinone-induced apoptosis also occurred in vivo, tunel assay was performed to determine apoptosis in tumor graft in mice. As indicated by brown spots, cells underwent apoptosis after kurarinone treatment (Figures 7A,B). To further confirm our results, we detected the expression of related proteins in tumor tissues. The results showed that kurarinone can up-regulate pro-apoptosis proteins cleaved-caspase-3, cleaved-caspase-9, Bax, and down-regulate antiapoptosis Bcl-2 expression (Figures 7C–E).