Tanshinone IIA combined with adriamycin inhibited malignant biological behaviors of NSCLC A549 cell line in a synergistic way

The human NSCLC cell line A549 and PC9, and the Human Lung Fibroblast (HLF) cell line were supplied by the Cell Bank of the Chinese Academy of Sciences (Shanghai, China) and cultured in RPMI1640 (Gibco, Carlsbad, CA, USA), supplemented with 10% fetal bovine serum (Gibco, Carlsbad, CA, USA) in a humidified incubator at 37 °C, 5% CO₂ atmosphere. Tanshinone IIA (Fig. 1a) was purchased from Sigma-Aldrich (St. Louis, MO, U.S.A.) and prepared as a 10 mM stock solution in dimethylsulfoxide (DMSO) (St. Louis, MO, USA). The solution was serially diluted in a RPMI 1640 medium immediately prior to the experiments. ADM (Fig. 1b) was purchased from Sigma-Aldrich and prepared as a 10 mM stock solution in normal saline (NS) which was serially diluted in RPMI 1640 medium immediately prior to experiments. Pancreatin, penicillin and streptomycin were purchased from Gibco (Invitrogen Life Technologies, Carlsbad, CA, USA). All the reagents were of analytical grade.

Cell proliferation was evaluated using the CCK8 (Dojindo Laboratories, 119 Kumamoto, Japan) according to manufacturer’s instructions. Briefly, A549, PC9 or HLF Cells (6 × 10³/90 μL/well) were plated into 96-well plates in triplicate and cultured for 24 h before onset of treatment. Then cells were treated with ADM, tanshinone IIA and combination of both drugs at a fixed molar ratio over a broad dose range to establish growth curves for 48 h. After that, cells were incubated for an additional 2 h with CCK-8 reagent (100 μl/mL medium). The absorbance was determined at 450 nm wavelength with a reference wavelength of 630 nm using a microplate reader (BioTek, Winooski, 126 VT, USA). The proliferative inhibition rate was measured using the Optical Density and calculated using the formula: proliferative inhibition rate = (1-treatment group/control group) × 100%. The IC₅₀ (50% inhibitory concentration) value was calculated by nonlinear regression analysis using GraphPad Prism software (San Diego, CA, USA).

The isobologram analysis for the combination study was based upon the Chou-Talalay method to determine combination indices (CI). The data obtained with the CCK8 assay was normalized to the vehicle control and expressed as % viability. Then, the data was converted to Fraction affected (Fa; range 0–1; where Fa = 0 represents 100% viability and Fa = 1 represents 0% viability) and analyzed with the CompuSyn™ program (Biosoft, Ferguson, MO) based upon the Chou and Talalay median effect principle [14, 15]. The CI values reflect the ways of interaction between two drugs. CI < 1 indicates synergism, CI = 1 indicates an additive effect, and CI > 1 indicates antagonism [16].

A549 cells (1 × 10⁶/1 mL/well) were plated in 6-well plates and allowed to adhere for 24 h. Confluent monolayer cells were scratched by a 200 μL pipette tip and then washed three times with 1 × PBS to clear cell debris and suspension cells. Fresh serum-free medium with different drug treatments were added, and the cells were allowed to close the wound for 48 h. Photographs (magnification, ×100) were taken at 0 h and 48 h at the same position of the wound. The migration distance was calculated by the change in wound size during the 48 h period using Adobe Photoshop CS6 software.

A549 cells (5 × 10⁴) were resuspended in 200 μl of serum-free medium containing different drug treatments and seeded on the top chamber of the 8 μm pore, 6.5 mm polycarbonate transwell filters (Corning, NY, USA), whose inserts were coated with a thin layer of 0.25 mg/ml Matrigel Basement Membrane Matrix (BD Biosciences, Bedford, MA). The full medium (600 μl) containing 10% FBS was added to the bottom chamber. The cells were allowed to migrate through the filters for 48 h at 37 °C in a humidified incubator with 5% CO₂. The cells attached to the lower surface of membrane were fixed in 4% paraformaldehyde at room temperature for 30 min and stained with 0.5% crystal violet. The cells on the upper surface of the filters were removed by wiping with a cotton swab. The number of stained cells on the lower surface of the filters was counted under the microscope (magnification, ×100). A total of 5 fields were counted for each transwell filter.

After incubation at 37 °C in an atmosphere of 5% CO₂ for 48 h, the treated cells were detached by trypsinization, collected, washed twice with cold PBS and fixed in 5 mL 75% cold ethanol at 4 °C for 24 h. The cells were again washed twice with PBS and incubated with 500 μl RNase (50 μg/mL) for 30 min at 37 °C, and then labeled with propidium iodide (PI, 0.1 mg/mL) then incubated at room temperature in the dark for 30 min prior to analysis. For each measurement, at least 20,000 cells were counted. Cell cycle analysis was performed by analyzing PI staining levels by flow cytometry (Beckman Coulter, USA). Data was analyzed using ModFit (Verity Software House, Inc, Topsham, ME).

Cell apoptosis was determined by PI and Annexin V-FITC staining (KeyGEN Biotech, Nanjing, China). In brief, the treated cells were incubated for 48 h, washed twice with ice-cold PBS, the collected cells were then resuspended in 200 μl of binding buffer and incubated with 5 μl each of Annexin V-FITC and PI for 15 min in the dark at room temperature, according to the manufacturer’s instructions. The cells were analyzed immediately after staining, using a FACScan flow cytometer (Becton-Dickinson). For each measurement, at least 20,000 cells were counted.

Apoptosis was detected using the In Situ Cell Death Detection Kit (Roche Molecular Bioscience, Mannheim, Germany) following manufacturer’s instructions. Apoptotic cells were imaged using a fluorescence microscope (Olympus, Tokyo, Japan). For each sample, three photomicrographs of random fields were taken at 400× magnification, and cells were scored as apoptotic or viable and counted. The percentage of apoptotic cells was determined by counting the TUNEL-positive cells and dividing the number by the total number of cells.

Immunofluorescence assay was applied to detect the expression of Cleaved Caspase-3. The treated cells were washed with PBS and then fixed with 4% paraformaldehyde for 15 min at room temperature. Permeabilization was done with 0.3% Triton X-100 for 30 min and then blocked with 5% normal FBS for 1 h at room temperature. After that cells were incubated overnight at 4 °C with anti-Cleaved Caspase-3 (1:200, Cell Signaling Technology, Beverly, MA) primary antibody. Secondary anti-mouse (1:500, Alexafluor488, Invitrogen, Carlsbad, USA) antibody was added for 1 h at room temperature in the dark. After washing with PBS three times, the coverslips were mounted on slides by using mounting medium containing DAPI (Invitrogen) and observed using a fluorescence microscope (Olympus, Tokyo, Japan) (magnification, ×400).

Western blotting analysis was applied for the re-confirm via molecular biological method. All the selected proteins extracts of each group cells were resolved by 10% SDS-PAGE and transferred on PVDF (Millipore, Bedford, MA, USA) membranes. After blocking, the PVDF membranes were washed four times for 15 min with TBST at room temperature and incubated with primary antibodies. The following primary antibodies were used: anti-Bax, anti-Bcl-2, anti-Caspase-3, anti-Akt, anti-phospho-Akt, anti-PI3K, anti-phospho-PI3K (all 1:1000; Cell Signaling Technology, Danvers, MA, USA), anti-VEGF (1:1000; Abcam, Cambridge, MA, USA), anti-Cleaved Caspase-3 (1:500; Cell Signaling Technology, Danvers, MA, USA), anti-VEGFR2, (1:200; Cell Signaling Technology, Danvers, MA, USA) and anti-GAPDH (1:2000; Cell Signaling Technology, Beverly, MA). Following extensive washing, membranes were incubated with secondary horseradish peroxidase (HRP)-conjugated secondary antibodies (1:1000; Cell Signaling Technology, Danvers, MA, USA) for 1 h. After washing 4 times for 15 min with TBST at room temperature once more, the immunoreactivity was visualized by enhanced chemiluminescence (ECL kit, Millipore, Billerica, MA, USA), and membranes were exposed to KodakXAR-5 films (Sigma-Aldrich). Relative optical density (ROD, ratio to GAPDH) of each blot band was quantified by using National Institutes of Health (NIH) image software (Image J 1.36b).

To predict the possible interaction of small molecules and the selected proteins, Discovery Studio (DS) 2.5 (Accelrys Software Inc, San Diego, CA) was applied to the molecular docking algorithm in this study. The calculation of root mean square deviation (RMSD) was carried out for the validation of the veracity for the selection of molecular docking modules in DS 2.5. The three-dimensional (3D) crystal structures of proteins were selected from PDB (http://​www.​rcsb.​org/​pdb/​).​The 3D structure of tanshinone IIA was downloaded from The PubChem Project (http://​pubchem.​ncbi.​nlm.​nih.​gov/​) with a CID of 164676. The DS 2.5 was run on a localhost9943 server. The docking procedure includes the following steps. Firstly, the water molecules in the proteins were removed and the hydrogen atoms were added to the proteins. Secondly, small molecules and selected proteins were refined with CHARMM. Thirdly, the active sites of proteins were defined by two methods: according to internal ligand’s binding site and automatically with DS 2.5. Lastly, small molecules were docked into the active sites of the proteins with the appropriate parameter settings. Through a series of algorithms, 10 different orientations were randomly generated. Each orientation was subjected to simulated annealing molecular dynamics simulation. ADM simulation was run consisting of a heating phase from 300 to 700 K with 2000 steps, followed by a cooling phase back to 300 K with 5000 steps. The energy threshold for Van der Waals force was set at 300 K. We further refined the simulation result by running a short energy minimization, consisting of 50 steps of steepest descent followed by up to 200 steps of conjugate gradient using an energy tolerance of 0.001 kcal/mol.

All experiments were performed in triplicate and repeated at least three times, a representative experiment was selected for the figures. Data was presented as mean value ± standard error and was analyzed using SPSS 15.0 software by one-way ANOVA with Dunnett’s post hoc test and Turkey’s post hoc test for multi-group comparisons (except the IC₅₀ values which were calculated by nonlinear regression analysis using GraphPad Prism software.). Student’s t-test was used for paired data. A p value of 0.05 or less was considered as significant. The drug interactions were assessed using multiple effect analysis based on the Chou-Talalay method.

As shown in Fig. 2 and Additional file 1, both ADM and tanshinone IIA inhibited the proliferation of the tested cell lines in a time- and dose-dependent manner, with HLF cells showing a lowest IC₅₀ value of ADM and a highest IC₅₀ value of tanshinone IIA among the tested cells. These data hinted that HLF cells displayed a stronger sensitivity to ADM, and a weaker sensitivity to tanshinone IIA, compared with the NSCLC A549 cell line and the NSCLC PC9 cell line.

Guided by the IC₅₀ values determined for the single drugs, the combinations of the ADM and tanshinone IIA were evaluated at the 1:20 (ADM: tanshinone IIA) fixed molar ratio for 48 h. Compared to any individual drug, drug combination exerted a much stronger inhibition of cells growth, except HLF cells. In A549 cells, drug combination treatment had a synergistic inhibitory effect (CI <1) when Fa value was ≤0.67 (Fig. 2b). The synergism of drug combination treatment (CI <1) was observed when Fa value was ≤0.664 in PC9 cells (Fig. 2d). For HLF cells, ADM combined with tanshinone IIA induced significant antagonistic growth inhibition (CI>1) when Fa value was ≤0.99 (Fig. 2f) . The summary of CI value and the concentration of the separate drugs in combination at 50% Fa were shown in Table 1.

Since A549 cell line is broadly used in lung cancer research area, we choose it as the further research object. To identify a combination that achieved maximal biological function, the migration and invasion ability in A549 cells were investigated by wound healing assay and transwell assay. Figure 3 showed that the migration distances and the invasive cell numbers were significantly decreased after 48 h drug treatment. Meanwhile, the combined simultaneous treatment showed the least migration distance and the invasive cell number in the results.

After verifying the anti-proliferation effect of tanshinone IIA and ADM, the distribution of cell cycles were detected by Flow cytometry. As shown in Fig. 4a, the cell population in G1 phase was decreased in both single drug treatment groups and drug combination groups, with the latter showing more deduction. At the same time, tanshinone IIA group increased the S phase cell population, the ADM and drug combination groups increased the G2 phase cell percentage in comparison with the untreated cells.

Then, we detect the cell apoptosis via Flow cytometry. Both single drug treatment and drug combination increased the proportion of early (dots in the lower right quadrant) and late apoptosis (dots in the upper right quadrant) in A549 cells (Fig. 4b). However, there was no statistical significance among the tested groups. TUNEL assays were performed to detect the DNA fragmentation in A549 cells after different drug treatments. The presence of TUNEL-positive cells (stained green), showing occurrences of DNA strand breaks, also indicates apoptosis in cells. Quantification revealed that all the tanshinone IIA groups, ADM groups, and drug combination groups increased the TUNEL-positive cells (Fig. 4d). Taken together, co-treatment more potently induced apoptosis compared with single treatment in A549 cells.

In order to explore the involved signal pathway, we performed western blotting to measure the levels of VEGF, VEGFR2, PI3K, p-PI3K, Akt, p-Akt, Bcl-2, Bax, Caspase-3, and Cleaved Caspase-3 upon single drug treatment groups and drug combination groups in A549 cells. Results revealed that both single drug treatment and drug combination up-regulated Bax, Cleaved Caspase-3 expression levels, but down-regulated VEGF, VEGFR2, Bcl-2, Caspase-3, p-Akt, and p-PI3K expression levels, with the total Akt, PI3K, and GAPDH levels remaining the same (Fig. 5a, b, c, d). Immunofluorescence assays were performed and the results revealed that Cleaved Caspase-3 level was consistently elevated by both single drug treatment and drug combination treatment in A549 cells (Fig. 5e, f). The effect of drug combination groups showed significant difference compared with single drug treatment groups in both immunofluorescence assay and western blot assay.

Molecular docking algorithm was applied to predict the possible interaction of small molecules and the selected proteins. First of all, as shown in Table 2, the RMSD of Akt2 (PDB-ID: 2JDR), Bcl2 (PDB-ID: 4IEH), PI3K (PDB-ID: 4J6I), and VEGFR-2 (3VHE) were 0.6208 Å, 1.370 Å, 0.8333 Å, and 0.3568 Å, respectively, when CDOCKER module in DS 2.5 was applied in the algorithms. It presented the veracity of CDOCKER in the study.

Secondly, as shown in Figs. 6, 7, 8 and 9, tanshinone IIA could be docked into active sites of all the proteins with individual binding modes, when compared with ADM. Tanshinone IIA could form H-bonds with Lys181 and aromatic interactions with Phe163 in the endogenous ligand’s active site of Akt2 (PDB-ID: 2JDR), while ADM formed H-bonds with Leu158, Glu236, Lys277, Asp440, Asn280, Thr292, and Asp293, aromatic interactions with Phe163, Val166, and Met282 (Fig. 6). Tanshinone IIA could form H-bonds with Arg105 and aromatic interactions with Arg66 in the endogenous ligand’s active site of Bcl-2 (4IEH), while ADM formed H-bonds with Ala59, Arg66, Asn102, and Try161, H-bonds plus aromatic interactions with Gly104, and Arg105 (Fig.7). Tanshinone IIA could form H-bonds with Lys890 and aromatic interactions with Met953 in the endogenous ligand’s active site of PI3K (4J6I), while ADM could only form H-bonds with Val882, Ala885, Asp884, Thr887, Lys890, Asp950, and Met953 (Fig. 8). Only tanshinone IIA could form H-bonds with Cys919, aromatic interactions with Leu840 and Val848 in the endogenous ligand’s active site of VEGFR-2 (3VHE) (Fig. 9). Thus, these results indicated that tanshinone IIA may display the anti-tumor effect with similar molecular mechanisms of ADM: to interact with the proteins in the same active sites, but to interact with different residues.