Human pancreatic cell line PANC-1, MIA PaCa-2 and HPAF-II were purchased from Chinese Academy of Life Science (Shanghai, China). DMEM medium was purchased from HyClone Laboratories Inc. (Shrewsbury, NJ, USA). Fetal bovine serum (FBS) was from Gibco Co. (Carlsbad, CA, USA). Baicalein was purchased from Xinran Co. (Shanghai, China). Gemcitabine and docetaxel were purchased from Sigma-Aldrich (St. Louis, MO, USA) and were dissolved according to the manufacturer's instructions. DMSO and DAPI were from Roche.

PANC-1, MIA PaCa-2 and HPAF-II cells were cultured in DMEM medium containing 10% FBS and 1% penicillin and streptomycin in incubator with 5% CO₂ at 37°C. Drugs at different dosage were given at the indicated time point.

Drug sensitivity was detected using the MTT assay. Briefly, cells were trypsinized and seeded in 96-well plates (Corning Inc., Corning, NY, USA) at 5×10³ cells per well. The cells were cultured overnight and then replenished with fresh medium containing drugs at indicated concentrations. The cells were then incubated for 48 h. A total of 20 μl of MTT (Sigma-Aldrich) dissolved in PBS at 5 mg/ml was added directly to the wells at the indicated time points. The plates were then incubated for an additional 4 h at 37°C for MTT reaction. The supernatant was then removed. A total of 100 μl of DMSO was added to dissolve the formed formazan crystals, and the optical density was measured at 490 nm on a PerkinElmer 2030 VICTOR X Multilabel Plate Reader (Perkin-Elmer, Waltham, MA, USA). The results are represented as the average value of 3 independent experiments. The percentage of viable cells was calculated as cell viability (%) = (OD of treatment/OD of control) × 100.

The viability of PANC-1, MIA PaCa-2 and HPAF-II cells were examined by MTT assay as described above and the CI values were analyzed using the Calcusyn software, which calculates CI value by the following equation: CI = (D)1/(Dx)1 + (D)2/(Dx)2 + (D)1(D)2/(Dx)1(Dx)2, where (Dx)1 and (Dx)2 are the doses for x% inhibition by drug 1 and drug 2 alone. (D)1 and (D)2 are the doses in combination that inhibit cell growth by x%. A CI value of 1 indicates additive effects of the two agents, while a CI value greater than 1 indicates antagonism effects, and less than 1 indicates synergism effects.

Viable PANC-1 cells were plated in 6-well plates for 24 h, followed by treatment with indicated drugs. After 48 h, cells were fixed with 4% paraformaldehyde for 20 min, followed by DAPI staining for 10 min in the dark. Finally, cells were detected using immunofluorescence microscopy (DSY5000X, OPPNO, Beijing, China).

Cell Light™ EdU kit was purchased from RiboBio Co., Ltd. (Guangzhou, China) and the experiment was conducted according to the manufacturer's instructions. Briefly, prepared 50 μM EdU DMEM medium solutions were added to drug treated PANC-1 cells in 96-well plates, followed by incubation for 2 h and thereafter washing with PBS, 4% paraformaldehyde was used to fix the cells for 30 min, and then 2 mg/ml glycine was given to neutralize the remaining paraformaldehyde. Apollo^® staining reaction solution was used to incubate PANC-1 cells in the dark for 30 min, and then washed with 0.5% TritonX-100 PBS 3 times. Finally, Hoechst 33342 was added for 30 min and images were taken from immunofluorescence microscopy (DSY5000X, OPPNO).

PANC-1 cells were seeded on 6-well plates (Corning Inc.). After incubation for 24 h, each well was initiated by scratching with a sterile 10 μl pipette tip, followed by washing with PBS three times, and treated with indicated drugs at 37°C. After 24 h, the distance between cell edges was analyzed using the ImageJ software.

After treatment with drugs, PANC-1 cells were digested by 0.25% trypsin from 6-well plates, and then collected and incubated with 70% ethanol overnight. After fixation, cells were stained with propidium iodide (PI) for 30 min at room temperature. The cells were washed with ice cold PBS 3 times before loaded to the flow cytometer (FACS Calibur, BD BioSciences). The results were then analyzed with ModFit software according to the manufacturer's instructions.

PANC-1 cell samples were digested by 0.25% trypsin and centrifuged at 1500 rpm for 5 min and fixed overnight in 2.5% glutaraldehyde at 4°C. Then the samples were fixed in 1% osmium acid, dehydrated, and placed in embedding molds in a standard fashion. Appropriate areas were selected and ultrathin sections of 0.08 μm were stained with lead citrate and uranyl acetate. Those sections were then examined with a transmission electron microscope (JEM1230, JEOL, Ltd., Tokyo, Japan).

PANC-1 cells were treated with drugs for 48 h and lysed in RIPA buffer, followed by denaturation. Protein concentration was measured by bicinchoninic acid assay system (Beyotime, Shanghai, China). Protein samples were separated by SDS-PAGE gel and electrophoretically transferred to nitrocellulose membranes. The membranes were blocked with 5% non-fat milk for 30 min and incubated with anti-caspase-3, anti-cleaved-caspase-3, anti-PARP and anti-cleaved-PARP antibodies (1:1,000; Cell Signaling Technology, Danvers, MA, USA) at 4°C overnight. The membranes were then incubated with goat anti-rabbit/anti-mouse secondary antibody conjugated with horseradish peroxidase (1:3,000; Cell Signaling Technology) and then membranes detected using an enhanced chemiluminescence detection kit (Thermo Scientific). β-actin was used as the internal control.

For statistical analyses, Prism 5 software was used. Statistical analyses were performed using the Student's t-test. A P-value of <0.05 was considered to indicate a statistically significant difference.

We determined the morphologic changes of PANC-1, MIA PaCa-2 and HPAF-II cells after treatment with baicalein (200 μM) by inverted microscope. After treatment, cells exhibited a more distinct cell profile and more shrunken morphology, while DMSO-treated control group showed a blurred cell profile and adhered tightly to the well (Fig. 1A–C). Then, we detected the effect of baicalein on cell viability in vitro using MTT assay. Consistently, cell viability of baicalein-treated PANC-1 was significantly suppressed (Fig. 1D). As shown in Fig. 1D, 200 μM baicalein triggered significant decrease in cell viability. Notably, the suppressive effects of baicalein on cell viability of PANC-1, MIA PaCa-2 and HPAF-II cells was in a dose-dependent manner.

To determine whether baicalein has a synergistic effect with gemcitabine/docetaxel, we administered baicalein in combination with different concentrations of gemcitabine/docetaxel to PANC-1. As shown in Fig. 2A, treatment with gemcitabine (2 μM) mildly affected cell viability of PANC-1 with 82.03±5.033%. When it was combined with baicalein (10 μM), cell viability decreased to 36.5±2.848%. Similarly, baicalein (10 μM) plus gemcitabine (5 or 10 μM) remarkably affected cell viability compared with gemcitabine (5 and 10 μM) alone (27.77±1.302% vs. 61.97±4.756%, P<0.01; 18.67±1.65% vs. 54.73±4.033%, P<0.01). The CI values calculated by Calcusyn software were 0.191, 0.204 and 0.179 when baicalein and gemcitabine were concurrently administered with the ratios of 5:1, 2:1 and 1:1, respectively (Fig. 2C). Similar to gemcitabine, docetaxel also exhibited synergistic effect with baicalein. Treatment with baicalein (10 μM) in combination with docetaxel (2, 5, and 10 nM) resulted in remarkable decrease in cell viability when compared with docetaxel alone (2, 5, and 10 nM) (39.17±2.109% vs. 64.2±2.969%, P<0.01; 31.97±1.525% vs. 54.07±2.282%, P<0.01; 20.7±1.858% vs. 46.83±1.742%, P<0.001) (Fig. 2B). The CI values were 0.201, 0.192 and 0.099, when baicalein (μM) and docetaxel (nM) were administered with the ratios of 5:1, 2:1 and 1:1, respectively (Fig. 2D). Consistently, similar results were obtained from MIA PaCa-2 and HPAF-II cells treatment with baicalein in combination with gemcitabine/docetaxel (Fig. 2E–L), suggesting that baicalein has synergistic effects with gemcitabine/docetaxel on suppressing cell viability of PANC-1, MIA PaCa-2 and HPAF-II cells.

It is well-known that gemcitabine and docetaxel could inhibit proliferation of pancreatic cancer cells. However, whether baicalein has synergistic effects with gemcitabine/docetaxel on proliferation of pancreatic cancer cells is unknown. Therefore, we used EdU assay to determine the effects of cell proliferation by combination treatment. Relative proliferation ability was normalized to DMSO treated group by the ratios of proliferating cells which were stained red in the EdU assay. The relative proliferation ability of baicalein (10 μM) treated group was lower than control, but no statistical significance (Fig. 3A and B). The relative proliferation ability of PANC-1 cells treated with baicalein (10 μM) in combination with gemcitabine (2 and 10 μM) showed significantly decreased proliferation when compared to gemcitabine alone (2 and 10 μM) (2.677±2.677% vs. 17.48±3.243%, P<0.05; 1.496±0.8111% vs. 4.72±0.469%, P<0.05) (Fig. 3A and B). Similarly, the relative proliferation ability of cells treated with baicalein (10 μM) in combination with docetaxel (2 and 10 nM) was significantly inhibited compared with docetaxel alone (2 and 10 nM) (29.36±5.822% vs. 45.11±8.995%; 9.051±3.642% vs. 23.91±3.953%), but no statistical significance (Fig. 3A and B). Our data suggest that baicalein might be a potential adjuvant to strengthen the anti-proliferation effects of the first-line clinical drugs on pancreatic cancer cells.

Cell cycle blockade usually leads to growth inhibition of cancer cells. To investigate the effects of baicalein alone and combination treatment on cell cycle of PANC-1 cells, we conducted flow cytometry to analyze the proportion of different phases in cell cycle. The results showed that baicalein alone (50 and 100 μM) induced remarkable cell cycle arrest in the S-phase (42.95 and 45.74%) compared with control (34.22%) (Fig. 4A and C). Combination treatment with baicalein (10 μM) and gemcitabine (2 μM) resulted in a significant obvious change than gemcitabine (2 μM) alone (21.97 vs. 6.46%). Similar results were observed between combination treatment with docetaxel (2 nM) plus baicalein (10 μM) and docetaxel (2 nM) alone (Fig. 4B and D). Therefore, baicalein might curb cell growth of PANC-1 by blocking the cell cycle.

To further determine the effects of baicalein on cell migration, we used wound healing assay to assess the migration ability of PANC-1 cells. As expected, inhibitory effects of baicalein on migration of PANC-1 cells was in a dose-dependent manner. As shown in Fig. 5A and B, either 10 or 20 μM baicalein alone exhibited no significant effects. However, baicalein at 50 μM induced statistically significant reduction in recovery ratio of PANC-1 cells compared with control (20±3.225% vs. 45.6±0.665%; P<0.01). Next, we investigated whether baicalein could reinforce the inhibitory ability of gemcitabine/docetaxel on migration of PANC-1 cells in vitro by using scratching assay. Treatment with baicalein (10 μM) in combination with gemcitabine (2 μM) showed more stronger ability of inhibiting cell migration compared with gemcitabine alone (2 μM) (6.45±1.91% vs. 27.85±3.90%; P<0.01). Similarly, combination of baicalein (10 μM) and gemcitabine (10 μM) exhibited significantly stronger suppressive effects on PANC-1 cells when compared with gemcitabine alone (10 μM) (4.30±1.73% vs. 25.85±3.72%; P<0.05). Baicalein (10 μM) plus docetaxel (10 nM) exhibited more obvious inhibitory effects than docetaxel alone (10 nM) on migration of PANC-1 cells (0.33±0.57% vs. 12.67±0.96%; P<0.01). These results indicate that baicalein could enhance the capacity of gemcitabine/docetaxel to inhibit migration of PANC-1 cells.

We further investigated whether baicalein-induced suppression of cell growth of PANC-1 cells is associated with cell apoptosis, we performed DAPI staining after treatment with drugs for 48 h. As expected, treatment with baicalein at 10 μM induced increased apoptosis in PANC-1 cells, compared with control (1.922±0.551% vs. 0.377±0.132%, P<0.05; Fig. 6B). Then we further performed TEM to confirm the above results. Consistently, apoptotic change in PANC-1 cells was found after treatment with baicalein alone at 50 μM (Fig. 6A). These data suggest that baicalein treatment could induce apoptosis of PANC-1 cells in vitro. To further uncover the molecular mechanism of apoptosis induced by baicalein, we performed western blot to detect the expression level of caspase-3, cleaved-caspase-3, PARP and cleaved-PARP, which are considered as classical apoptosis-related molecules. The results showed that, after treatment with baicalein, the expression level of caspase-3 was significantly decreased, accompanied by increased expression of cleaved-caspase-3. Immunoblotting for total PARP protein showed no significant change after baicalein treatment, while protein level of cleaved-PARP was markedly increased (Fig. 6E). Our data suggest that apoptosis-related caspase-3/PARP signaling pathway might play an important role in baicalein-induced apoptosis of PANC-1 cells.

Next, we further investigated whether baicalein could enhance pro-apoptotic effects of gemcitabine/docetaxel on PANC-1 cells using DAPI staining. Interestingly, treatment with 10 μM baicalein in combination with 2 nM docetaxel resulted in significantly increased apoptosis of PNCA-1 cells compared with 2 nM docetaxel alone (75.86±5.184% vs. 15.13±2.169%; P<0.001). Similar results were observed at high concentrations of docetaxel. Docetaxel at 10 nM triggered significant apoptosis of PNCA-1 cells with the apoptosis rate of 76.07±9.61%, while 10 nM docetaxel together with 10 μM baicalein induced more severe cell apoptosis of 87.63±5.19% (P<0.05) (Fig. 6B and D). In addition, TEM showed that apoptosis changes existed in all drug treated groups (Fig. 6A). These results demonstrate that baicalein has a potential to enhance pro-apoptotic effects of docetaxel on PANC-1 cells in vitro. However, treatment with 10 μM baicalein plus 2 μM gemcitabine did not induce significant change in apoptosis rate of PANC-1 cells when compared with 2 μM gemcitabine alone. Similarly, there was also no significant difference in apoptosis of PNCA-1 cells treated with 10 μM baicalein plus 10 μM gemcitabine compared with 10 μM gemcitabine alone (Fig. 6B and C). Moreover, caspase-3/PARP signaling pathway was significantly activated when treated with baicalein in combination with gemcitabine/docetaxel (Fig. 6F and G). These data suggest that baicalein might exert different effects on pro-apoptosis effect of different drugs and the underlying mechanism might be complicated.