PD markedly decreases PAC cells viability.

The cytotoxicity of 13 classical Chinese herbal formulae from Chinese medicine book “Shang Han Treatise” (A-M, listed in Additional file 1: Table S1) on PAC cells was evaluated using the CCK-8 assay. As shown in Additional file 1: Fig. S2, after 72 h exposure of PAC cells to the 13 formulae at concentration of 100 μg/mL, certain formulae such as PD (A), Yinchenhao Tang Decoction (C) and Wuzhuyu Decoction (H), showed prominent cytotoxicity. Among them, PD (A) demonstrated the strongest cytotoxicity on the two PAC cell lines, Gemcitabine HCl (GEM), a clinically used anti-PAC agent, was set as a positive control.

To examine the inhibitory action of PD on PAC cells, two PAC cell lines were exposed to PD at different concentrations for a period of 24 h, 48 h and 72 h. After a 48 h exposure to PD, the half-maximal inhibitory concentration (IC50) values were 130.61 μg/mL ± 27.74 for MIA PaCa-2 cells and 119.62 μg/mL ± 11.59 for BxPC-3 cells. Whereas IC50 for 72 h treatment were 128.92 μg/mL ± 15.25 and 99.84 μg/mL ± 16.36, respectively. These data showed that PD dose- and time-dependently decreased the viability of PAC cells (Fig. 1A, B).

Fig. 1

PD induces cell cycle arrest at G2/M phase and mitochondrial apoptosis in PAC cells. The viability of PAC cells (A, B) after PD treatment at concentrations of 0, 75, 100, 150 and 200 μg/mL was determined using CCK-8 for 24 h, 48 h and 72 h, respectively. Representative images of cell cycle distribution of PAC cells (C, D) after treated with PD (0–200 μg/mL) for 12–48 h, respectively. Quantification data of cell cycle distribution for PAC cells (E, F) are presented. The levels of G2/M phase-related regulatory proteins (p-CDC25C (Ser216), CDC25C, Cyclin B1, p-CDK1(Tyr15), CDK1 and p-Histone H3 (Ser10)) in PAC cells (G, H) were analyzed using immunoblotting following exposure to PD at concentrations of 0, 50, 100 and 200 μg/mL for 12 h. The expression levels were normalized to GAPDH. Double staining with Annexin V-FITC and PI was conducted to evaluate apoptosis in PAC cells (I, J) following PD treatment for 48 h (0–200 μg/mL) and then quantified (right panel). Apoptosis-associated protein expression was determined using immunoblotting in MIA PaCa-2 (K, L) and BxPC-3 (M, N) following PD treatment at concentrations of 0, 50, 100 and 200 μg/mL for duration of 12, 24 and 48 h. The expression levels were normalized to β-actin. 0.5 μM of GEM was chosen as the positive control. Results are presented as mean ± S.D. of triplicate independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001 compared with the control

PD triggers PAC cells to arrest at G2/M phase and undergo apoptosis.

To investigate the effects of PD on cell cycle progression, PAC cells were treated with PD at different concentrations. After 12–48 h, the cells were then harvested and analyzed with PI-staining flow cytometry. Following 12 h exposure of 200 μg/mL PD, there was a notable increase in the number of cells in the G2/M phase, rising from 20% to 66.1% in MIA PaCa-2 cells and 20.4% to 42.3% in BxPC-3 cells (Fig. 1C–F). CDC25C phosphatase dephosphorylates CDK1, then activated cyclin B1-CDK1 complex drives cell progression through G2/M checkpoint and enter mitosis, accompanied with increased p-Histone H3 (Ser10) expression [22, 23]. After 12 h treatment of PD, the expression levels of G2/M phase-related proteins, such as p-CDC25C (Ser216), CDC25C and p-CDK1 (Tyr15), were decreased. Conversely, a dose-dependent upregulation of cyclin B1 and p-Histone H3(Ser10) protein levels was observed (Fig. 1G, H). These results suggested that PD arrested cell cycle at G2/M phase and possibly at M phase.

Notably, the accumulation of SubG1 phase cells were markedly enhanced following 24 and 48 h of PD treatment (Fig. 1C–F). Hence, we performed Annexin V/PI double staining assay to evaluate whether PD induced cell apoptosis. As shown in Fig. 1I and J, a dose-dependent induction of apoptosis was observed in PAC cells after 48 h of PD treatment, with increments were from 4.9% to 95.1%, and from 7.2% to 97.6%, respectively. The immunoblotting results verified that PD treatment resulted in a decrease in the expression of anti-apoptotic protein Bcl-2 and an increase in the expression of cleaved PARP, cleaved caspase 3, and cleaved caspase 9 in a dose- and time-dependent manner (Fig. 1K–N). Overall, the data suggest that PD triggers cell cycle arrest at G2/M phase and promotes apoptosis.

BTW from PD exerts the predominant anti-PAC effect.

To identify which herb that predominates the inhibitory effect of PD on PAC cells, the herbal content of PD were proportionately separated according to their original weight ratio in PD (BTW:HL:HB:QP = 2.5:2:2:1). The separated herbal contents were then variedly combined producing 15 dissembled groups(Table 1). The cytotoxicity of 15 disassembled groups on PAC cells was evaluated using CCK-8 assay after 72 h. As shown in Fig. 2A, B, BTW (group 2) showed the strongest cytotoxicity against two cell lines, with IC50 value of 27.9 μg/mL ± 1.25 in MIA PaCa-2 cells and 30.24 μg/mL ± 4.45 in BxPC-3 cells, which was about fourfold lower than PD. Additionally, all the dissembled groups that contain BTW (group 6, 7, 9, 10, 11 and 12) showed greater cytotoxicity with lower IC50 value than PD, suggesting that BTW is the most effective ingredient contributing to the anti-pancreatic effect of PD. Therefore, CCK-8 assay was used to evaluate the inhibitory effect of BTW at different concentrations during a range of time periods in PAC cells (Fig. 2C, D). Our findings indicate that BTW exerts a dose- and time-dependent inhibitory effect on the growth of PAC cells.

Table 1 15 disassembled groups from PD

Herbal content of PD were proportionately separated according to their original weight ratio in PD (BTW:HL:HB:QP = 2.5:2:2:1), producing 15 dissembled groups. PD, Pulsatilla Decoction; BTW (Bai Tou Weng), Pulsatillae chinensis; HB (Huang Bai), Phellodendron chinense; HL (Huang Lian), Coptis chinensis; QP (Qin Pi), Cortex fraxini.

Fig. 2

The cytotoxicity of 15 disassembled groups from PD on PAC cells. The IC50 values of 15 disassembled groups on PAC cells (A, B) at 72 h. The inhibitory effect on PAC cells (C, D) after treated with BTW at concentrations of 0, 25, 37.5, 50, 75 and 100 μg/mL for a period time of 24, 48 and 72 h. Data are presented as mean ± S.D. of triplicate independent experiments. *p < 0.05, ***p < 0.001 compared with the control

To examine whether the underlying anti-pancreatic mechanism of BTW was consistent with PD, we investigated the effects of BTW on inducing cell cycle arrest at G2/M phase and cellular apoptosis. In PAC cells following 12 h of treatment, BTW dose-dependently expanded the cell population in G2/M phase, with the proportion rising from 26.9% to 46.9% in MIA PaCa-2 cells and from 26.8% to 39.8% in BxPC-3 cells, respectively (Fig. 3A–D). Additionally, BTW reduced the protein levels of p-CDC25C (Ser216), CDC25C and p-CDK1 (Tyr15), while increased the protein expression of cyclin B1 and p-Histone H3 (Ser10) in the both cell lines (Fig. 3E, F). Moreover, BTW treatment for 24 h and 48 h remarkably accumulated PAC cells in the SubG1 phase (24.2% and 43.5% of MIA PaCa-2 cells, 28.5% and 36.4% of BxPC-3 cells) (Fig. 3A–D). Annexin V/PI staining assay detected that BTW dose-dependently caused a marked increase in apoptotic cells (Fig. 3G, H). The immunoblotting results consistently demonstrated that BTW dose- and time-dependently reduced Bcl-2 protein expression as well as activated PARP, caspase 3, and caspase 9, in both cell lines (Fig. 3I–L). Notably, treatment of either 40 μg/ml BTW or 200 μg/ml PD for 48 h started to stimulate the cleavage of PARP, caspase 3, and caspase 9 proteins, indicating that the antitumor action of 40 μg/ml BTW is comparable to 200 μg/ml PD. Collectively, the induction of cell cycle arrest at G2/M phase and caspase-dependent apoptosis in PAC cells by BTW shares similarities with the effects observed with PD, further supporting that BTW is the main active anti-pancreatic components in PD.

Fig. 3

BTW induces G2/M cell cycle arrest and activates mitochondrial apoptosis in PAC cells. Representative images of cell cycle distribution of PAC cells (A, B) following BTW treatment at concentrations of 0, 10, 20 and 40 μg/mL for duration of 12, 24 and 48 h, respectively. Quantification data of cell cycle distribution of PAC cells (C, D). The expression of the indicated proteins in PAC cells (E, F) was analyzed using immunoblotting after treatment with 0–40 μg/mL of BTW for 12 h. The expression levels were normalized to β-actin. Double staining of Annexin V-FITC and PI was conducted to detect apoptotic cell death in PAC cells (G, H) following BTW treatment (0–40 μg/mL) for 48 h. Quantification analysis was shown on the right panel. The expression of apoptosis-related proteins was evaluated using immunoblotting in MIA PaCa-2 (I, J) and BxPC-3 (K, L) cells following BTW treatment. β-actin was used as an internal reference for protein expression normalization. 0.5 μM of GEM was chosen as the positive control. Results are presented as mean ± S.D. of triplicate independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001 compared with the control

β-peltatin, a small molecule extracted from BTW, shows strong anti-PAC actions by triggering G2/M phase arrest and apoptosis.

We then made an effort to identify the key chemical responsible for BTW action (a separate manuscript under review). Briefly, the aim compound showed same molecular ions with podophyllotoxin and picropodophyllotoxin, but the retention time and the major fragment ions were different from the two reference standards, suggesting the aim compound was an isomer of these two standards. LC-TOF–MS analysis of both commercial BTW extract and original BTW herb confirmed the presence of this isomer. Analysis using 1H-NMR and 13C-NMR revealed the aim compound as β-peltatin with molecular formula C22H22O8 and molecular weight 414.4.

Next, we exposed PAC cells to the indicated concentrations of β-peltatin for a period of 24 h, 48 h and 72 h (Fig. 4A, B). The IC50 value of β-peltatin (72 h treatment) was 2.09 nM ± 0.72 in MIA PaCa-2 and 1.49 nM ± 0.37 in BxPC-3 (Table 2), indicating its pronounced contribution to the cytotoxicity of BTW and PD. To verify β-peltatin is the main compound of BTW and PD that exerted cell cycle arrest at G2/M phase and apoptotic cell death induction on PAC cells, we evaluated the cell cycle distribution and the mechanism of cell death. The results in Fig. 4C–F showed that 12 h treatment of 2 nM β-peltatin significantly induced G2/M phase arrest. Annexin V/PI double staining results reflected that 4 nM β-peltatin induced about 40–50% PAC cells underwent apoptosis (Fig. 4I, J). Consistently, β-peltatin dose- and time-dependently modulated G2/M phase-related pathway (Fig. 4G, H) and mitochondria-mediated caspase-dependent pathway (Fig. 4K–N). The data collectively indicate the dominant role of β-peltatin in the anti-PAC activity of BTW and PD by promoting cell cycle arrest at G2/M phase and cellular apoptosis.

Fig. 4

β-peltatin inhibits the growth of PAC cells by inducing G2/M cell cycle arrest and mitochondrial apoptosis. Cell viability of PAC cells (A, B) was examined following treatment with the indicated concentrations of β-peltatin for 24 h, 48 h and 72 h. Representative images of cell cycle distribution of PAC cells (C, D) following administration of 0–4 nM β-peltatin for 12–48 h, respectively. Quantification data of cell cycle distribution of PAC cells (E, F). The expression of G2/M phase-related proteins in PAC cells (G, H) were examined using immunoblotting following treatment with β-peltatin at concentrations of 0, 1, 2 and 4 nM for 12 h. The expression levels were normalized to GAPDH. Following 48 h of β-peltatin administration, the apoptotic PAC cells were assessed by Annexin V-FITC and PI-stained flow cytometry (I, J) and quantified (right panel). The expression levels of apoptosis-related proteins were evaluated by immunoblotting in MIA PaCa-2 (K, L) and BxPC-3 (M, N) cells with β-peltatin treatment at the indicated intervals and doses. The expression levels were normalized to β-actin. 0.5 μM of GEM was chosen as the positive control. Results are presented as mean ± S.D. of triplicate independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001, compared with the control

Table 2 IC50 values of β-peltatin and podophyllotoxin in PAC cells

The values of IC50 were determined using GraphPad Prism 5 to analyze the results of the CCK-8 assay (Fig. 4A, B, Additional file 1: Fig. S3A, B). The IC50 values were obtained from the mean ± S.D. of triplicate experiments.

β-peltatin is the isomer of podophyllotoxin, an aryltetralin-type lignan isolated from species of Podophyllum, which exhibited potent anticancer actions but clinically obsoleted due to its serious toxicity [24]. We therefore compared the cytotoxicity of β-peltatin and podophyllotoxin on PAC cells by using the CCK8 assay (Fig. 4A, B, Additional file 1: Fig. S3A, B). The IC50 value of podophyllotoxin (11.33 nM ± 0.69 in MIA PaCa-2 cells, 13.71 nM ± 0.31 in BxPC-3 cells) were about 5- to 9-fold higher than that of β-peltatin (2.09 nM ± 0.72 in MIA PaCa-2 cells, 1.49 nM ± 0.37 in BxPC-3 cells) (Table 2). Next, we compared the acute toxicity of β-peltatin and podophyllotoxin in mice. Results showed that the overall survival rate of mice after a single administration of 60 mg/kg β-peltatin and 60 mg/kg podophyllotoxin was 90% and 60%, respectively (Additional file 1: Fig. S3C), despite no significant alterations in body weight of survived mice in either β-peltatin or podophyllotoxin-treated groups (Additional file 1: Fig. S3D). Therefore, these data support that comparing with podophyllotoxin, β-peltatin has stronger PAC cell proliferation inhibitory activity and lower acute toxicity in mice.

β-peltatin suppresses tumor growth of BxPC-3 cells in vivo

We used the subcutaneously-xenografted BxPC-3 cells model to examine in vivo anti-tumor effect of β-peltatin. Intraperitoneal administration of 15 mg/kg β-peltatin (once weekly) significantly suppressed tumor growth (Fig. 5A) without any notable alterations in the body weight of mice compared to control group (Fig. 5B) or gross anatomy of primary organs (Fig. 5C). Notably, when compared with its isomer podophyllotoxin, β-peltatin inhibited tumor growth more effectively. Moreover, β-peltatin prolonged the survival time of mice by 65.85% (34 days) and podophyllotoxin by 31.71% (27 days) compared to controls (21 days) (Fig. 5D). H&E and immunohistochemical staining of resected tumors from β-peltatin-treated mice displayed sparse cellularity and negative Ki-67 expression (Fig. 5E, F). Furthermore, β-peltatin treatment notably elevated the levels of cleaved caspase 3 and p-Histone H3 (Ser10) proteins in comparison to the vehicle control group (Fig. 5G, H), which is consistent with the in vitro results. These findings indicate that β-peltatin has the potential to suppress pancreatic tumor growth in vivo, reduces Ki-67 expression, and elevates the expression of cleaved caspase 3 and p-Histone H3 (Ser10) in tumor tissues.

Fig. 5

β-peltatin inhibits pancreatic tumor growth of BxPC-3 cells in vivo. BxPC-3 cells were administered via subcutaneous injection into the right flank of nude mice. Upon the tumor reaching a size of 100 mm3, mice were randomly divided into three different groups: vehicle, β-peltatin (15 mg/kg) and podophyllotoxin (15 mg/kg). Measurements of tumor volume (A) and mice body weight (B) were recorded every other day. *compared with vehicle group; #statistical significance between β-peltatin and podophyllotoxin group. The mice were sacrificed when tumor volume reached 1000 mm3, the primary organs (C) were excised and photographed. Survival curves (D) and quantification data (right panel) for mice treated with vehicle, β-peltatin, and podophyllotoxin. Immunohistochemical analysis was conducted on paraffin-embedded tumor tissues using H&E staining, and antibodies against Ki-67, cleaved caspase 3 and p-Histone H3 (Ser10); E–H showed representative images (E, × 400) and quantification data (F–H). Results are presented as mean ± S.D. *p < 0.05, **p < 0.01, ***p < 0.001 compared with the control; #p < 0.05, ###p < 0.001 compared with the indicated group