Paeoniflorin and paeonol were purchased from Sigma Pharmaceutical Co. (Monticello, IA) for use as standards. The agents were accurately weighed and dissolved and serially diluted in 70% methanol to give concentrations of 150, 75, 37.5, 18.75, and 9.375 μg/ml. Calibration curves were plotted following linear regression of the peak areas. Ten grams of Cortex Moutan (Sun Ten Pharmaceutical Co., Ltd., Taichung, Taiwan) was combined with 20 ml of double-distilled water in a 50 ml centrifuge tube and agitated at 4˚C for 48 h. The resulting solution was filtered using 0.22 mm PVDF membrane, and the filtrate was freeze-dried to obtain 1.10 g of powder. The dried powder (46 mg) was combined with 1 ml of 70% methanol and placed in an ultrasonic bath for 30 min. Following centrifugation at 10,000 rpm for 10 min, the supernatant was filtered through a 0.45 mm filter and analyzed by HPLC. The HPLC system (Agilent 1100) comprised a G1310A isocratic pump with solvent cabinet, aG1328A manual injector, a G1314A variable wavelength detector (VWD) with standard flow cell (10 mm path length, 14 µl volume, 40-bar maximum pressure), and G2220AA 2D-Value Solution ChemStation software (Agilent Technologies, Waldbronn, Germany). The HPLC conditions for Cortex Moutan analysis were as follows. Analysis was performed with a ZORBAX Eclipse XDB-C8 column (4.6 × 150 mm, 5 μm, Supelco; Agilent Technologies). The mobile phase comprised 0.03% phosphate in water-acetonitrile (0 min, 90:10; 10 min, 75:25; 20 min, 60:40; 30 min, 40:60; 32 min, 85:15; 35 min, 90:10). The flow rate was 0.6 ml/min, and 10 ml portions were injected into the column. The column temperature was set to 20˚C. The detector detected the compounds in Cortex Moutan at the wavelength of 230 nm, and the retention times of paeoniflorin and paeonol were 11.6 and 27 min, respectively.

Cortex Moutan which was validated by HPLC was dissolved in double-distilled water and filtered using a 0.22- micron filter. The concentration of the stock solution was thenconfirmed by weighing after lyophilization. Mitomycin C (Kyowa Hakko Kirin Co., Tokyo, Japan), doxorubicin (Actavis Italy S.P.A., Milan, Italy), and cisplatin (ABIC Biological Laboratories Ltd., Netanya, Israel) were dissolved in normal saline buffer at 1 mg/ml to provide stock solutions, which were then diluted with cell culture medium to desired concentrations ranging from 0.0025 to 0.08 mg/ml. Human bladder papillary transitional cell line BFTC 905 and human bladder carcinoma cell line TSGH 8301 (Food Industry Research and Development Institute, Taiwan), as well asprimary normal human urothelial cell line HUC 4449 (ScienCell Research Laboratories, Carlsbad, CA), were used as cell models. BFTC 905 and TSGH 8301 cells were cultured in RPMI 1640 supplemented with 10% fetal bovine serum (FBS), 100 units/ml penicillin, and 100 μg/ml streptomycin. HUC 4449 cells were cultured in urothelial cell medium (ScienCell Research Laboratory, Carlsbad, CA) supplemented with 10% FBS, 100 units/ml penicillin, and 100 μg/ml streptomycin. All cell lines were cultured at 37˚C under a humidified atmosphere containing 5% CO₂, and all treatment with drugs was also performed under these conditions.

To assess the cytotoxicity of Cortex Moutan, cell viability was determined with the use of Cell Counting Kit-8 (CCK-8; Sigma-Aldrich Chemie GmbH, Steinheim, Germany). The reagent in the assay kit is converted by dehydrogenases in the mitochondria of live cells to generate a colored product that can be measured by absorbance at 450/650 nm, allowing the absorbance reading to be directly correlated with the number of viable cells. This assay was therefore used to analyze the viability of Cortex Moutan-treated cells relative to that of control cells whose absorbance value was taken to be 100%. The CCK-8 assay was performed as follows. BFTC 905, TSGH 8301, and HUC 4449 cells were seeded in 96-well culture plates at 1 × 10⁴ cells per well, cultured for 24 h, and then washed with phosphate-buffered saline (PBS) and starved of serum by culturing in 100 μl per well of culture medium without FBS for 24 h. The culture medium was removed, and fresh medium containing Cortex Moutan, cisplatin, doxorubicin, or mitomycin C at various concentrations was added to the wells in triplicate and allowed to incubate with the cells for 24 h. Cells in the control wells were incubated with fresh medium without drugs instead. Drug-treated and control cells were washed with PBS, and 10 μl of CCK-8 reagent was added per well and incubated with the cells at 37˚C and 5% CO₂ for 60 min before the culture plates were measured for absorbance at 450/650 nm in a microplate reader (Model 680; Bio-Rad, Hercules, CA). For each drug, cell viability at each drug concentration was calculated as follows:

The viability values were plotted against drug concentration, and the drug concentration at which 50% of the treated cells were nonviable (IC50) was estimated from the data. To estimate the IC50 we plotted x-y and fitted the data with a straight line (linear regression). IC50 value was then estimated using the fitted line, i.e.,

In addition, the selectivity index (SI) of each drug, representing the cytotoxic selectivity of the drug against cancer cells relative to normal cells, was calculated as follows:

Bladder cancer cells were seeded in 6-well culture plates at 1 - 2 × 10⁵ cells per well and cultured for 24 h, starved for serum in 2 ml per well of culture medium without FBS for 24 h, and then treated in triplicate with Cortex Moutan at various concentrations in culture medium for 24 h. Afterwards, the cells were completely dissociated by treatment with trypsin-EDTA, which was directly added at 2X concentration at 2 ml per well, and the cells were then dehydrated and fixed as single-cell suspensions in methanol at 2 ml per well at 4˚C for 24 h. The fixed cells were collected by centrifugation at 1500 rpm for 10 min, rehydrated and washed with ice-cold PBS, and stained by incubating with 10 mg/ml RNase A and 1 mg/mlpropidium iodide in PBS in the dark at room temperature for 30 min before being subjected to flow cytometry analysis for cell cycle status. Ten thousand cells of each sample type were collected for the analysis.

Bladder cancer cells were cultured for 24 h and treated in triplicate with various concentrations of Cortex Moutan for 24 h as described above and then dissociated to single cells with Trypsin-EDTA, directly added at 2X concentration at 1 ml per well of 6-well plates. By using components of the Annexin V-FITC Apoptosis Detection Kit (BioVision, Mountain View, CA), the trypsinized cells harvested from each well were resuspended in 500 μl of binding buffer, mixed with 5 μl of annexin V-FITC and 5 μl of 50 ug/mlpropidium iodide (optional), and allowed to incubate in the dark for 5 min before being subjected to flow cytometry analysis to detect apoptotic cells. Ten thousand cells of each sample type were collected for the analysis.

Bladder cancer cells (BFTC 905 and TSGH 8301) were seeded in 6-well culture plates at 1 - 2 × 10⁵ cells per well and cultured for 24 h, starved for serum in 2 ml per well of culture medium without FBS for 24 h, and then treated with 1 mg/ml Cortex Moutan in culture medium for 16 or 24 h. Cell total proteins were extracted from the treated cells and identified using a Bio-Rad protein assay (Bio-Rad, Hercules, CA) with bovine serum albumin (BSA) as a standard. Each lysate (10 µg) was resolved on denaturing polyacrylamide gels and transferred electrophoretically to a PVDF transfer membrane. The membranes were blocked with blocking buffer (Bio-Rad, Hercules, CA) for 1 h, incubated with primary antibodies at 4˚C overnight, and then incubated with horseradish peroxidase-conjugated secondary antibodies at room temperature for 1 h, with membrane washing carried out after each of the steps in TBST buffer. Antibody-bound proteins on the blots were detected with ECL Western Blotting Detection Reagents (GE Healthcare Biosciences, Piscataway, NJ), followed by chemiluminescent imaging using the BioSpectrum 800 system (UVP, Upland, CA).

Analysis of cell viability, cell cycle, or apoptosis at each drug concentration was performed with triplicate samples. Data are expressed relative to no-treatment controls as mean ± standard deviation (SD). The significance of differences between drug treatments was analyzed by one-way ANOVA followed by Dunnett’s post hoc test for multiple comparisons. A P-value < 0.01 was regarded as statistically significant.

The bioactive components of Cortex Moutan are paeoniflorin and paeonol. The distinctly different structures of these two compounds produced different solubility properties, which led to different HPLC retention times as detected at the 230 nm absorption wavelength. We ran calibration standards containing paeoniflorin and paeonol and found the retention times to be 11.6 and 27 min, respectively (Figure 1). Accordingly, we used the identical HPLC method to determine the presence of paeoniflorin and paeonol in our Cortex Moutan sample by retention time. As shown in Figure 2, the chromatogram displays a peak at 11.6 min and at 27 min, indicating that the Cortex Moutan sample contained both paeoniflorin and paeonol analytes.

To examine the cytotoxic activity of Cortex Moutan, we cultured two human bladder cancer cell lines in the presence of various concentrations of Cortex Moutan for 24 h and then analyze the viability of the cells by performing the CCK-8 assay. The results showed that the viability of human bladder cancer cells decreased with increasing concentration of Cortex Moutan (Figure 3). Cortex Moutan treatment reduced cell viability by as much as 74% and 82% in BFTC 905 and TSGF 8301 cells, respectively (Figure 3), and the Cortex Moutan concentration that corresponded to a 50% reduction in cell viability (IC₅₀) was 0.4633 and 0.5782 mg/ml, respectively. IC₅₀ values in BFTC 905 and TSGH 8301 cells were also determined for three common chemotherapeutic agents, cisplatin (0.044 and 0.0041 mg/ml, respectively), doxorubicin (0.0021 and 0.0019 mg/ml), and mitomycin C (0.0041 and 0.004 mg/ml) (Table 1). In order to test the cytotoxicity of Cortex Moutan in human urothelial cells, the same drug treatments and viability analysis were performed with primary normal human urothelial cells of the HUC 4449 line. We found that the viability of HUC 4449 cells were not significantly affected by treatment with Cortex Moutan at concentrations up to 1 mg/ml, but a reduced viability was observed at 2 mg/ml (Figure 3), giving an IC₅₀ value of 2.81 mg/ml (Table 1). In contrast, cisplatin, doxorubicin, and mitomycin C significantly reduced the viability of HUC 4449 cells at much lower concentrations and had respective IC₅₀ values of 0.0232, 0.0054, and 0.0102 mg/ml (Table 1). Taken together, the above data show that Cortex Moutan exhibited cytotoxic activity against human bladder cancer cell lines while being substantially less cytotoxic toward normal urothelial (HUC 4449) cells than mitomycin C, doxorubicin, and cisplatin.

To assess the selectivity of Cortex Moutan’s cytotoxicity against cancer cells rather than normal cells, we calculated the selectivity index (SI) of each drug for BFTC 905 and TSGH 8301 cells. The SI is defined as the ratio between the IC₅₀ in normal cells and the IC₅₀ in cancer cells for a given drug. A drug with a larger selectivity

^*Drug concentration at which 50% of the treated cells are nonviable relative to untreated cells. Data are shown as mean ± SD of three replicates.

index would exert a more potent cytotoxic effect on cancer cells with less cytotoxicity to normal cells, and thus has more selective cytotoxicity toward cancer cells. We found the SI values of Cortex Moutan for BFTC 905 and TSGH 8301 cells to be 6.30 and 5.05, respectively (Table 2), both far greater than 3, indicating Cortex Moutan’s highly selective cytotoxicity toward cancer cells [29] . In contrast, the SI values of mitomycin C, doxorubicin, and cisplatin for BFTC 905 cells and of doxorubicin and mitomycin C for TSGH 8301 cells were all less than 3 (Table 2), indicating these drugs’ relatively low selective cytotoxicity toward cancer cells. These results demonstrate that relative to mitomycin C, doxorubicin, and cisplatin, Cortex Moutan more effectively inhibited the growth of BFTC 905 and TSGH 8301 human bladder cancer cells under conditions that preserved the viability of normal cells.

Next, we examined the effect of Cortex Moutan on cell cycle progression in human bladder cancer cells. As shown by the results of flow cytometry analysis, treatment with Cortex Moutan at 0.25, 0.5, or 0.75 mg/ml significantly decreased the proportion of BFTC 905 cells in the G0/G1 phase and increased the proportion in the S phase (Figure 4(a)). In identically treated TSGH 8301 cells, the proportion in the G0/G1 phase also significantly decreased, but the proportion in the G2/M phase significantly increased (Figure 4(b)). These results indicate that Cortex Moutan induced cell cycle arrest in BFTC 905 cells at the G1/S transition, as well as in TSGH 8301 cells at the G2/M transition. Also, we observed marked increases in the sub-G1 populations of BFTC 905 and TSGH 8301 cells treated with 0.5 or 0.75 mg/ml Cortex Moutan (Figure 4), suggesting that Cortex Moutan induced cell damage in these cell lines, resulting in an increased occurrence of DNA aneuploidy and cell death.

As a number of studies have shown that arrested cell cycle progression may cause cells to undergo apoptosis [30]

^*A value >3 indicates high selectivity.

[31] , we wished to ascertain whether the Cortex Moutan—induced cell death observed was mainly due to apoptosis or necrosis. Through flow cytometry analysis, we found that nearly 7.5% of BFTC 905 cells were apoptotic in the absence of Cortex Moutan, and treatment with Cortex Moutan at 0.75 mg/ml significantly increased the percentages of both apoptotic cells and necrotic cells (Figure 5(a)). In TSGH 8301 cells, Cortex Moutan treatment induced primarily apoptotic death and secondarily necrotic death, both in a dose-dependent manner (Figure 5(b)). These results show that Cortex Moutan induced both apoptotic and necrotic death in bladder cancer cells,

with different cancer cell lines exhibiting different sensitivities to the cell death—inducing effects of Cortex Moutan treatment.

In tumor angiogenesis, angiogenic factors such as bFGF, VEGF, and IL-8 are produced by cancer cells to stimulate the formation of new blood vessels. Thus, inhibiting cancer cells’ expression of these angiogenic factors will decrease vascular growth, thereby inhibiting tumor growth, progression, and metastasis. We therefore further examined the effect of Cortex Moutan on the protein expression of bFGF, VEGF, and IL-8 in bladder cancer cells. As detected by western blot analysis, treatment with Cortex Moutan at 1 mg/ml for 16 or 24 h significantly decreased the level of VEGF in BFTC 905 and TSGH 8301 cells (Figure 6(b)).However, we detected a mild decrease in the levels of bFGF and IL-8 in BFTC 905 and TSGH 8301 cells (Figure 6(a) & Figure 6(c)). These results show that Cortex Moutan inhibited the protein expression of VEGF and slightly decreased the protein expression of bFGF and IL-8 in cancer cells.