Cinchonine induces apoptosis of HeLa and A549 cells through targeting TRAF6

HeLa and A549 cells were provided from Tianjin International joint Academy of Biomedicine, Normal human dermal fibroblast (NHDF) cells were provided from Professor Jun Dai (Tianjin University, China). RPMI-1640, DMEM, and FBS were purchased from Corning, Australia. Cinchonine (Xiensi Biochemical Technology Co, Tianjin, China) was dissolved in DMSO (Sigma-Aldrich). Phosphatase inhibitor and 3-(4,5-dimethyl-2-thiazoyl)-2,5-diphenyl-2H-tetrazolium bromide (MTT) were obtained from Sigma, USA. HEPES was purchased from Solarbio (Beijing, China). Apoptosis Detection Kit was purchased from Becton-Dickinson, USA. Polyvinylidene Fluoride (PVDF) membrane was purchased from Merk Millipore, USA. In Situ Apoptosis Detection Kit, POD was purchased from BOSTER (Wuhan, China). Protein A Sepharose beads were brought from Pierce, USA. Balb/c-nude mice and Kunming mice were purchased from Yi Sheng Yuan Bio Logical Technology Co. Tianjin, China.

Antibodies used in this study include the following: Ubiquitination antibody and rabbit anti-TRAF6 polyclonal antibody (Santa Cruz, USA); Rabbit polyclonal antibody against total AKT, phosphorylation-AKT (Thr-308), phosphorylation-AKT (Ser-473), total TAK1, phosphorylation-TAK1 (Thr-184/Thr-187), Bax, and Bcl-2 (Cell Signaling Technology, USA); Rabbit polyclonal antibody against TRAF6, ALEXA FLUOR 647 conjugated (Bioss Inc. USA); Rabbit anti β-actin polyclonal antibody, Goat anti-mouse, and goat anti-rabbit IgG polyclonal antibody (Beijing Zhongshan Golden Bridge Biotechnology Co., Ltd, China).

HeLa and A549 cells were cultured in RPMI-1640, while NHDF cells were cultured in DMEM, both media were supplemented with 10% FBS. All cells were incubated at 37 °C in a humidified atmosphere of 5% CO₂.

To first identify small molecules that could bind with the RING domain of TRAF6, the three-dimensional structure of the RING and ZINC finger domains (residues 50–159) of TRAF6 was obtained from the Protein Data Bank database with an access code of 3HCT [29]. The structure was used in the docking experiments, and the docking grid was placed over the RING domain (residues 67–124) of TRAF6, which is believed to interact with Ubc13. We carried out computational screening of a small library of chemical compounds established by Jkchemical Sigma, and Alfa Aesar that would include 1792 commercially available compounds. All structures were built using a 2D/3D editor-sketcher and were minimized to a local energy minimum using the CHARMm-like force field. This library was docked into TRAF6 implemented by AutoDock4.10 software. Compounds were filtered by the predicted binding free energy. Further filtering rules were guided by chemical diversity, rigidity, novelty and actual commercial availability.

Cells were seeded into 96-well plate at density of 4.1 × 10³ cells/well. Different concentrations (1, 5, 10, 50, 100, and 250 μM) of cinchonine were added to these wells. Cis-platinum (16.7 μM) was used as a positive control. After treating with cinchonine for 48, 72, and 96 h respectively, the medium with different concentrations were removed and cells were incubated with 20 μL MTT (5 mg/mL) for 4 h at 37 °C. The MTT solution was discarded and formazan was dissolved in 150 μL DMSO. The absorbance of each well was read using a Microplate reader (Shanghai Kehua, China) at 490 nm [30]. Cell inhibition ratio was calculated using the equation:

Half maximal inhibitory concentration (IC₅₀) was calculated by regression curve.

Apoptosis of the cells were determined using the Annexin V-FITC Apoptosis Detection Kit (Becton-Dickinson, Franklin Lakes, NJ, USA) according to the manufacturer’s protocol [30]. In brief, cells were seeded into 60-mm plate and the confluence was allowed to reach 50%–60%. After which, according to MTT result, cinchonine was added at concentrations of 100 μM, 150 μM, and 180 μM for HeLa cells, and at concentrations of 150 μM, 180 μM, and 220 μM for A549 cells at 37 °C for 48 h. Whereafter, cells were washed twice in cold PBS and then resuspended in 1× binging buffer, and 100 μL of the cell suspension was transferred to a 5 mL culture tube, to which 5 μL Annexin V-FITC and PI were added. The mixture was incubated for 15 min at room temperature in the dark. Results were immediately analyzed with a FACScalibur flow cytometer and cell Quest Pro 5.1 (BD, Biosciences, Franklin Lakes, NJ, USA).

Cells were extracted with lysis buffer containing 50 mM HEPES (pH 7.4), 150 mM NaCl, 1% NP-40. The lysate was centrifuged at 12000 × g for 20 min at 4 °C, and the supernatant containing total protein was harvested. Subsequently, the total protein was exposed to anti-body against AKT, which was immobilized on protein A beads (Pierce). The resulting beads were incubated at 4 °C overnight with gentle rotation, washed 3 times with lysis buffer, and separated at 800 × g for 3 min. The desired protein bands were visualized by Western blotting [31].

When the confluence reached 60%–70%, cells were treated with 180 μM of cinchonine for different time. After which, cells were treated with 20 μg/mL of lipopolysaccharide (LPS) for 3 h to induce activation of AKT [31]. To induce activation of TAK1, 1 μg/mL of LPS was used to treat the cells for 15 min [32]. The protein were extracted with lysis buffer for 30 min on ice. Extracts were centrifuged at 12000 × g for 20 min at 4 °C, and the supernatants containing total protein were harvested. Each sample containing 50 μg protein was separated by 10% SDS-PAGE and transferred to PVDF membranes. The membrane was blocked in 5% non-fat milk and incubated overnight at 4 °C with anti-bodies against AKT (1:1000), phosphorylation-AKT (1:1000), TAK1 (1:1000), phosphorylation-TAK1 (1:1000), ubiquitin (1:1000), Bax (1:1000), Bcl-2 (1:1000), β-actin (1:1000). Then the membranes were incubated at room temperature for 1 h with their corresponding secondary anti-bodies. The membranes were incubated with ECL solution (CWBIO) and exposed to the film (FUJI, Japan). The membranes were stripped and reblotted with β-actin antibody to verify the equal loading of protein in each lane. Image J software v.1.48u (National Institutes of Health, Bethesda, MD, USA) was used to quantify the intensity of protein bands.

Animals were maintained in Institute of Radiation Medicine Chinese Academy of Medical Science. All experimental procedures using these mice were in accordance with requirements and guidelines for treating experimental animals approved by the National Institutes for Food and Drug Control of China. All animal experiments have been approved by The Ethics Committee on animal experiments of Institute of Radiation Medicine Chinese Academy of Medical Science (The approval number is: (2016) 1–008). Balb/c-nude mice (four weeks old and gender random) were first subcutaneously inoculated with cancer cells (5 × 10⁶ cell/mL) to establish transplanted mode of cervical cancer. After feeding the nude mice for 3 weeks post implantation, the tumor volume was approximately 586.84 mm³ when intratumorally injection commenced. The inoculated mice were then divided into: experimental group-I to be intratumorally injected with cinchonine (180 μM), experimental group-II to be intratumorally injected with cinchonine (360 μM), and control group (treated with medium). The injection volume is 100 μL. Dosage of cinchonine administrated to nude mice are 0.265 mg/Kg (180 μM) and 0.530 mg/Kg (360 μM). The weight of nude mice and the shortest and longest diameter of tumor were measured with calipers with an interval of every other day. Tumor volume (mm³) was calculated using the following standard formula: (the longest diameter) × (the shortest diameter)²/2. The hyperplasia rate was calculated using: (“the volume of nude mice in day 2, 4, 6, 8, 10, 12, and 14” – “the volume of nude mice in day 2”)/“the volume of nude mice in day 2”. There were 5 nude mice in each group.

For the acute toxicity test, we used high-dosage intravenous injection, 500 μL of cinchonine (360 μM) with injected into 5 male and 5 female Kunming mice, and injected mice were observed for 14 d.

After treating with cinchonine for 14 d, paraffin sections were prepared from tumor tissues in the following manner. The resulting paraffin sections were then treated with an Apoptosis Detection kit (BOSTER, Wuhan, China) according to the manufacturer's protocol [33]. Briefly, paraffin sections were first incubated in electric dry oven at 60 °C for 30 min before they were placed into dimethyl benzene for 10 min and in fresh dimethyl benzene for another 10 min. Subsequently, these section were placed in a gradient of ethanol (5 min each ethanol absolute, 95 and 75% ethanol). The proteinase K was used to incubate them for 15 min at room temperature before being washed with 0.01 M phosphate buffer saline solution. After these paraffin sections were cultivated in terminal deoxynucleotidyl transferase (TdT) buffer containing TdT and biotinylated DNA uracil nucleoside triphosphate (dUTP) in TdT buffer, they were incubated in a moist atmosphere at 37 °C for 2 h and washed with TBS. Paraffin sections were then sealed and cultivated with biotin in a wet atmosphere for 30 min at room temperature before being washed with TBS. These sections were then incubated with Strept Avidin-Biotin Complex (SABC) for 30 min in 37 °C and washed with TBS. Lastly, paraffin sections were dyed with diaminobenzidine (DAB) and counterstained with Hematoxylin [34, 35], and the final toxicity analysis was viewed under an optical microscope.

The fluorescent probe, 7-(N,N-diethylamino) coumarin-N-(4-bromobenzyl)-3-carboxamide, was conjugated to the terminal olefin of cinchonine according to the following procedure. A mixture of 7-aminocoumarin (150.0 mg, 0.35 mmol), cinchonine (51.4 mg, 0.18 mmol), Pd(OAc)₂ (2.00 mg, 8.7 μmol), PPh₃ (4.80 mg, 18.0 μmol) and Et₃N (49.0 μL, 0.40 mmol) in anhydrous toluene (2 mL) was capped under N₂ atmosphere and heated at 110 °C for 24 h [36]. The resulting mixture was then allowed to cool to room temperature, filtered through Celite^TM, and washed with CHCl₃. The filtrate was concentrated to dryness under reduced pressure, and the resulting crude was purified using reversed-phase HPLC to give the desired coumarin probe conjugated cinchonine as yellow solid (17.0 mg, 15%).

Conditions for reversed-phase HPLC: Column, Cosmosil 5C18-AR300 (Nacalai Tesque, Inc.) 10 × 250 mm; Mobile phase A, 0.1% TFA in H₂O; B, 0.1% TFA in CH3CN; Gradient elution, 0–4 min at 5% B, 4–34 min at 5–95% B, 34–35 min at 95% B; Flow rate at 4 mL/min; UV detection at 254 nm. 1H NMR (400 MHz, CDCl₃, 25 °C) δ 9.42 (t, J = 5.8 Hz, 1H, NH), 9.10 (d, J = 5.5 Hz, 1H), 8.70 (s, 1H), 8.40 (dd, J = 14.8, 8.6 Hz, 2H), 8.26 (d, J = 5.4 Hz, 1H), 7.99 (t, J = 7.7 Hz, 1H), 7.84 (t, J = 7.7 Hz, 1H), 7.44 (d, J = 9.0 Hz, 1H), 7.34 (q, J = 9.5 Hz, 4H), 6.67 (dd, J = 9.0, 2.4 Hz, 1H), 6.59 (s, 1H), 6.53 (d, J = 15.8 Hz, 1H), 6.49 (d, J = 2.3 Hz, 1H), 6.33 (dd, J = 15.8, 7.7 Hz, 1H), 4.63 (d, J = 5.6 Hz, 2H), 4.40 (dd, J = 14.0, 7.1 Hz, 1H), 3.64–3.62 (m, 1H), 3.50–3.44 (m, 6H), 3.27 (q, J = 10.3 Hz, 1H), 2.79 (q, J = 8.2 Hz, 1H), 2.48 (t, J = 11.6 Hz, 1H), 2.14 (s, 1H), 2.02 (t, J = 8.4 Hz, 1H), 1.82 (q, J = 10.5 Hz, 1H), 1.24 (t, J = 7.1 Hz, 6H), 1.15–1.08 (m, 1H); ESI-HRMS m/z calcd for C₄₀H₄₃N₄O₄ ([M + H]+) 643.3279, found 643.3282.

Cells grown on glass coverslips were incubated with the fluorescent probe conjugated cinchonine for 1 h and washed with ice-cold PBS. These cells were subsequently fixed in 95% ethanol at room temperature for 30 min and permeabilized with 0.2% Triton X-100. After being blocked with 5% BSA in PBS, the cells were incubated with Rabbit polyclonal antibody against TRAF6, ALEXA FLUOR 647 conjugated (Bioss Inc. USA) for 2 h at 37 °C [37]. Co-localization of modified cinchonine and TRAF6 was analyzed using fluorescence microscope. The fluorescence microscope used for immunofluorescence observation is Nikon eclipse 80i and the light is Nikon INTENSILIGHT C-HGFI. The analysis software is NIS-Elements. The image is a single layer. There was no quantitative analysis of the co-localization.

Statistical analysis was performed with SPSS software and the experimental values which between difference groups were shown as the mean ± standard deviations. The statistical significance of differences between the control and treated groups were determined by Student’s t-test. Values *p < 0.05 and **p < 0.01 are considered significant.

We found cinchonine (Additional file 1: Figure S1), a member of Cinchona alkaloid family, to be the most effective in binding to the RING domain of TRAF6. More specifically, we found four possible areas for binding of cinchonine to the RING domain. As shown in Fig. 1a, after analyzing all four possible areas and comparing their free energies, the most suitable binding mode for cinchonine was determined, as it possesses the lowest free energy (△G = −6.64 kcal/mol) of the four areas analyzed. In this binding mode, cinchonine appears to be surrounded by amino acid residues of Asp-57, Glu-59, Phe-60, Pro-63, Leu-64, Leu-74, and Ala-76, thereby revealing also significant interactions with residues preceding (54–66) the RING domain (67–124). From the expanded view as shown in Fig. 1b, there appear to be hydrophobic interactions between the quinoline motif of cinchonine and residues Phe-60, Leu-64, and Ala-76 of TRAF6. More significantly, we were able to verify that the hydroxy group (−OH) of cinchonine forges a key hydrogen bond with Asp-57, whereas the terminal olefin of cinchonine does not interact with amino acid residues of TRAF6. These structural observations provide a critical basis to pursue possible visualization of such interaction in cells using a synthetically modified cinchonine that would be substituted with a fluorescent probe. In particular, we envisioned that if an appropriate fluorophore could be conjugated at the terminal olefin of cinchonine, it would remain outside the RING domain and would not interfere the binding of cinchonine at this region.

Western blot assay showed that HeLa and A549 cells established high expression levels of TRAF6 (Additional file 1: Figure S2). In Figs. 2a and b, cinchonine inhibited HeLa and A549 cells proliferation in a dose- and time-dependent manner. In NHDF cells, using the same concentration gradient, the inhibition ratio was significantly lower than HeLa and A549 cells (Fig. 2c). With the concentration of cinchonine being set at 180 μM, inhibition rates were 16.08, 22.07 and 24.95%, respectively, for durations of 48, 72, and 96 h. We subsequently calculated the half maximal (50%) inhibitory concentration (IC₅₀) value to be 180 μM through the computer-estimated using the Pharmacologic Calculation Program [38] and employed this concentration for the following experiments.

After treatment with cinchonine for 48 h, the percent of early apoptosis in control group was 4.53% ± 0.58. The percent of apoptosis increased after being treated with cinchonine. The fraction of early apoptosis increased to 20.92% ± 1.21 when treated with 100 μM of cinchonine and to 31.03% ± 1.87 with 150 μM of cinchonine, while reaching 38.69% ± 1.62 with 180 μM of cinchonine. As a positive control, Cis-platinum (16.7 μM) showed 33.69% ± 4.10 of cells in the state of early apoptosis (Fig. 3a). Therefore, we have demonstrated here that cinchonine could induce early apoptosis in HeLa cells as well as in A549 cells (Fig. 3b).

To understand the underlying molecular mechanism of cinchonine induced apoptosis, ubiquitination and phosphorylation levels of AKT were analyzed in these two cell lines. Immunoprecipitation showed that ubiquitination of AKT was down-regulated after being treated with cinchonine for 1, 2 and 3 h (Fig. 4a). Furthermore, in Fig. 4b, phosphorylations of AKT at Thr-308 and Ser-473 were significantly impeded after treatment for 3 and 6 h. Phosphorylation levels of TAK1 at Thr-184 and Thr-187 were analyzed and found reduced after exposing these two cell lines to cinchonine (Fig. 5). Levels of downstream apoptosis-related kinases protein Bax (pro-apoptotic protein)) and Bcl-2 (anti-apoptotic protein) were also examined. After cells were treated with cinchonine for 12, 24, 48, 72 and 96 h, the expression level of Bax was increased significantly, while the level of Bcl-2 decreased in a time-dependent manner in both cell lines (Fig. 6). These results provide tangible evidence that cinchonine could influence both AKT and TAK1 signaling pathways as well as their downstream proteins like though its binding with TRAF6.

Tumor model in nude mice was successfully established, and the tumorigenic rate was 100%. Uptake of food and drink of nude mice was normal, and the mental state was good. The tumor growth ratio of the experiment group-I was 2.267 ± 0.467, while the experiment group-II was 1.324 ± 0.367, which is significantly lower than the control group: 2.839 ± 0.583 (Fig. 7a). Weights of these mice were approximately 18–22 g (Additional file 2: Table S1). Uptake of food and drink of all Kunming mice was normal and the mental state was good without any death or loss of mobility after treatment with cinchonine for 14 d.

Comparing tumor tissues of nude mice from the three groups, in situ TUNEL assay shows that experimental groups have more nuclei being dyed than that of the control group, thereby revealing the ability of cinchonine to induce apoptosis of transplanted tumor cells in nude mice (Fig. 7b).

Results of our docking study suggest that cinchonine could effectively bind with TRAF6. To visualize such interaction in cells, a coumarin-derived fluorescent probe was conjugated specifically to the terminal olefin of cinchonine via synthesis (Fig. 8a). Immunofluorescence staining experiment using this modified cinchonine was performed. Modified cinchonine (green fluorescence) and TRAF6 (red fluorescence) were seen co-located in the cytoplasm of HeLa and A549 cells (Fig. 8b), thereby supporting the assessment that cinchonine could bind with TRAF6 in cells.