Chemopreventive and Chemotherapeutic Effects of Intravesical Silibinin against Bladder Cancer by Acting on Mitochondria Free

Bladder cancer poses a health problem worldwide. There were 52,810 new cases of bladder cancer occurring in men and 18,170 in women in the United States in 2009 (1). Superficial bladder cancer accounts for approximately 70% of all bladder cancer cases. The standard treatment for patients with superficial bladder cancer is transurethral resection (TUR) of tumors (2). However, approximately 60% to 70% of these tumors will recur, with 25% showing progression to a higher stage or grade (3). Intravesical chemotherapy is widely used as an adjuvant therapy to prevent recurrence and progression of superficial disease after TUR (4). Although many chemical agents have shown some evidence of activity (4), their toxicity and incomplete efficacy have limited their use as common intravesical agents. These factors highlight the urgent need to search for novel adjuvant intravesical agents.

Silibinin, a natural flavonoid, is the major bioactive component of silymarin isolated from milk thistle (Fig. 1A). Accumulating evidence indicates that silibinin has anticancer activity in various tumor cells, including cancers of prostate (5–9), breast (10, 11), skin (12), colon (13), lung (14), kidney (15, 16), and bladder (17–20). There is increasing interest in elucidating the mechanisms of action for silibinin as an effective agent for chemoprevention and chemotherapy against various types of cancers. Induction of apoptosis is believed to be one of the major mechanisms of action for silibinin against cancer cells (21), although the details are yet to be elucidated. There are multiple signaling pathways controlling the fate of cells undergoing apoptosis and determining their sensitivity to apoptosis inducers. Mitochondria play a central role in apoptosis, undergoing a number of profound changes early within the apoptotic cascade (22). The loss of mitochondrial membrane potential (ΔΨ_m) releases cytochrome c (Cyto c) into the cytosol, which subsequently forms a complex with apaf-1 and pro-caspase-9 for the activation of caspase-9, resulting in the final activation of caspase-dependent DNase that is responsible for DNA degradation (23). In addition, the apoptosis-inducing factor (AIF), the first caspase-independent cell death effector to be identified, translocates from the mitochondria via the cytosol to the nucleus, where it interacts with DNA and causes nuclear condensation and DNA fragmentation (24), features seen in apoptotic cells. However, the exact role of AIF in carcinogenesis and cancer therapy is not fully understood. For example, Gallego and colleagues showed that AIF determines the sensitivity of non–small-cell lung carcinomas to chemotherapy (25), and the chemoresistance of non–small-cell lung carcinoma is overcome through restoration of the AIF-dependent apoptotic pathway (26). In contrast, Urbano and colleagues reported that AIF suppresses chemical stress-induced apoptosis and maintains the transformed state of tumor cells, indicating that AIF may play an important role in tumorigenesis (27).

Several reports suggest that silibinin induces apoptosis through caspase activation in human bladder carcinoma RT4 (a well-differentiated papillary noninvasive tumor phenotype) cells (17, 19, 20). In addition, oral administration of silibinin has been reported to prevent N-butyl-N-(4-hydroxybutyl)nitrosamine (OH-BBN)-induced mouse bladder carcinogenesis (28), suggesting a value of silibinin in bladder cancer prevention and therapy. However, as the standard of care in patients with bladder cancer who have high-risk clinical and pathologic features, intravesical therapy receives more attention by urologists and their patients than oral administration (4). Therefore, the purpose of this study was to determine the effects of intravesical silibinin against superficial bladder cancer. In the present study, using in vitro and in vivo bladder cancer models, we also explored the molecular mechanisms of silibinin-induced apoptosis, focusing on 2 mitochondrial cell death pathways, namely the cyto c/caspase-dependent and the AIF/caspase-independent pathways.

The human bladder transitional cell carcinoma 5637 cell line was from the American Type Culture Collection (ATCC) in 2008. The cell line was used in fewer than 6 months after receipt or resuscitation. ATCC performs authentication of cell line via short tandem repeat profiling, karyotyping, and cyto c oxidase subunit I testing. Only mycoplasma tests were performed for the cell line authentication in our laboratory. We did not carry out additional testing to authenticate cell line, but its morphology and behavior were consistent with ATCC descriptions. Cells were cultured in RPMI-1640 supplemented with 10% of fetal bovine serum (Gibco) and 1% of penicillin–streptomycin at 37°C, in humidified air containing 5% of CO₂. Six-week-old athymic BALB/c nu/nu mice and 6- to 8-week-old female Sprague-Dawley rats were obtained from the Experimental Animal Center of Xi'an Jiaotong University. Animal care and protocols were approved by the Institutional Animal Care and Use Committee of Xi'an Jiaotong University; all the animal experiments were performed in adherence with the NIH Guidelines on the Use of Laboratory Animals.

Silibinin, MTT [3-(4,5-dimethylthiazol-2-yl)-2, 5-diphenyltetrazolium bromide], N-methyl-N-nitrosourea (MNU) and pan caspase inhibitor z-VAD-fmk were purchased from Sigma Chemical Co. Antibodies against procaspase-3, AIF, Histone H1, glyceraldehyde-3-phosphate dehydrogenase (GAPDH), cyto c, Smac (second mitochondria-derived activator of caspases)/DIABLO (direct IAP-binding protein with low pI), Omi/HtrA2 (human high temperature requirement protein A2), and survivin were from Santa Cruz Biotechnology, Inc. Antibodies against caspase-9, cleaved caspase-9, cleaved caspase-3, cleaved caspase-8, poly (ADP-ribose) polymerase (PARP), cyto c oxidase IV (COX IV), and peroxidase- or FITC-conjugated secondary antibodies were from Cell Signaling Technology, Inc. The enhanced chemiluminescence (ECL) detection system was obtained from Amersham.

Cell viability was assessed using a tetrazolium-based assay. Cells (5637) were incubated in the absence or presence of silibinin for the indicated times, and then washed once and incubated with 0.5 mg/mL of MTT at 37°C for 1 hour. The reagent was reduced by living cells to form an insoluble blue formazan product. After incubation, cells were lysed with DMSO. Colorimetric analysis using a 96-well microplate reader was performed at a wavelength of 490 nm. The experiments were performed in triplicate.

Cells (5637) were exposed to different doses of silibinin (50, 100, and 200 μM) and/or z-VAD-fmk for 48 hours. The cells were collected and subjected to annexin V and propidium iodide (PI) staining using an Annexin V-FITC Apoptosis Detection Kit, following the protocol provided by the manufacturer. Apoptotic cells were then analyzed by flow cytometry.

Caspase activity was assessed by a colorimetric system. Briefly, after treatment with desired doses of silibinin and/or z-VAD-fmk, cells were lysed in a lysis buffer by freeze/thaw cycles. Lysates were clarified by centrifugation at 10,000 g (4°C) for 10 minutes. One hundred micrograms of protein was incubated with 30 μL of caspase assay buffer and 2 μL of caspase-3 (DEVD-pNA) or caspase-9 (LEHD-pNA) colorimetric substrate at 37°C for 4 hours. The optical density of the reaction mixture was quantitated spectrophotometrically at a wavelength of 405 nm using a 96-well microplate reader.

Preparation of mitochondrial and cytosolic extracts has been previously described (29). Briefly, cells were incubated in permeabilization buffer [250 mmol/L of sucrose, 20 mmol/L of HEPES/KOH (pH of 7.4), 1 mmol/L of EGTA, 1 mmol/L of EDTA, 1 mmol/L of DTT, 0.1 mmol/L of PMSF, 1 μg/mL of chymostatin, 1 μg/mL of leupeptin, 1 μg/mL of antiparin, and 1 μg/mL of pepstatin A]. Cells were homogenized and centrifuged at 500 g (4°C) for 10 minutes to pelletize the nucleus and cell debris. The supernatants were further centrifuged at 13,000 g for 30 minutes. Cytosolic supernatants and mitochondrial pellets were collected. The pellets were then suspended in mitochondrial lysis buffer [150 mmol/L of NaCl, 50 mmol/L of Tris-HCl (pH of 7.4), 1% of NP-40, 0.25% of sodium deoxycholate, and 1 mmol/L of EGTA and protease inhibitor]. After 1 minute of vigorous vortex and 15 minutes of centrifugation at 15,000 g (4°C), the mitochondrial supernatants were collected.

Preparation of nuclear extracts was carried out as described previously (30) with some modifications. Briefly, cells were incubated with buffer A [10 mmol/L of HEPES/KOH (pH of 7.9), 10 mmol/L of KCl, 0.1 mmol/L of EDTA, 1 mmol/L of DTT and 0.4% of NP-40]. The lysates were placed on ice for 30 minutes and then centrifuged at 15,000 g (4°C) for 5 minutes. The pellet was resuspended in buffer B [20 mmol/L of HEPES (pH of 7.9), 400 mmol/L of NaCl, 1 mmol/L of EDTA, 1 mmol/L of DTT and 10% of glycerol] and incubated for 60 minutes on ice, and then further centrifuged at 15,000 g (4°C) for 15 minutes. The supernatants containing the nuclear fraction were collected. All the samples were stored at −80°C until use.

Cell lysates were prepared in lysis buffer [10 mmol/L of Tris-HCl (pH of 7.4), 150 mmol/L of NaCl, 0.1% of SDS, 1 mmol/L of EDTA, 1 mmol/L of EGTA, 0.3 mmol/L of PMSF, 0.2 mmol/L of sodium orthovanadate, 1% of NP-40, 10 mg/mL of leupeptin and 10 mg/mL of aprotinin]. For immunoblot analyses, 30- to 60-μg samples of protein (total lysates, mitochondrial, cytosolic, and nuclear fractions) were subjected to SDS-PAGE on 12% or 16% Tris-glycine gels, and separated proteins were transferred onto nitrocellulose membranes by Western blotting. The membranes were blocked with 5% milk for 1 hour at room temperature and probed with primary antibodies against desired molecules overnight at 4°C followed by peroxidase-conjugated secondary antibody for 1 hour at room temperature (25°C). The bands of proteins of interest were visualized using the ECL detection system followed by exposure to X-ray film. The relative intensity of each band was determined by using Glyko BandScan software.

Mitochondrial membrane potential collapse is a critical step that occurs in cells undergoing apoptosis (31). In the present study, ΔΨ_m was determined using JC-1 probes. Cells were harvested after 48 hours of exposure to DMSO or desired doses of silibinin and centrifuged at 400 g (4°C) for 5 minutes. The cell pellet was resuspended in 0.5 mL of JC-1 solution and incubated at 37°C for 20 minutes. After rinsing, the cells were analyzed by flow cytometry. A dot plot of red fluorescence (FL2, living cells with intact ΔΨ_m) versus green fluorescence (FL1, cells with lost ΔΨ_m) was recorded.

Cells were plated on slides overnight and then treated with 200 μM of silibinin and/or z-VAD-fmk for 48 hours. The cells were fixed in 4% paraformaldehyde for 25 minutes at room temperature. They were rinsed with PBS and blocked for 1 hour in 3% bovine serum albumin at room temperature, incubated with anti-AIF antibody (1:200) at 37°C for 2 hours. Then, the cells were incubated with secondary antibody (antigoat-FITC, 1:100) in the dark at 37°C for 50 minutes. Finally, stained cells were visualized under Olympus IX-50 fluorescence inverted microscope.

The cultured 5637 cells were detached by trypsinization, washed, and resuspended in serum-free RPMI-1640 medium. Cells (1 × 10⁶ cells in 100 μL) were then injected subcutaneously into the right flank of athymic BALB/c nu/nu mice to initiate tumor growth. After 7 days, 98% of mice grew visible tumors. Tumor-bearing mice were randomized to control and various treatment groups, and gavaged with the following preparations of silibinin in physiologic saline (0.9% NaCl): (a) the control group (n = 8), saline; (b) silibinin-100 (n = 10), 100 mg/kg/day of silibinin; (c) silibinin-200 (n = 10), 200 mg/kg/day of silibinin; (d) silibinin-300 (n = 10), 300 mg/kg/day of silibinin. All treatments were administered for 32 days. Body weight and diet consumption were recorded twice weekly throughout the study. The tumor volume was measured twice a week and was calculated by the formula 0.5236 ab², wherein a represents the long axis and b represents the short axis of the tumor. At the end of experiment, tumors were excised, weighed, and then fixed in 4% of paraformaldehyde for further analyses. The necropsies were carried out after all the mice were sacrificed. Gross and histologic survey was performed in liver, lung, spleen, and kidney.

N-methyl-N-nitrosourea-induced rat bladder tumor model utilized has been previously described with some modifications (32). Briefly, female Sprague-Dawley rats were anesthetized with intraperitoneal chloral hydrate. After the bladder was drained, 0.2 mL (2 mg) of MNU in saline was instilled intravesically via a 22-gauge Teflon angiocatheter every other week, for a total of 4 doses (weeks 2, 4, 6, and 8). The rats were divided randomly into 7 groups. Groups 1 to 3 (25 rats/group) received intravesical MNU every other week for 4 doses, and silibinin dissolved in DMSO (200 mg/kg for Group 1, MNU + silibinin-200, and 400 mg/kg for Group 2, MNU + silibinin-400) or DMSO alone (Group 3, vehicle control, MNU + DMSO) weekly beginning week 1 for 17 weeks. Group 4 (MNU, 25 rats) received intravesical MNU only. Group 5 (silibinin-400, 15 rats) received no MNU treatment, but 400 mg/kg of intravesical silibinin weekly for 17 weeks. Group 6 (DMSO + saline, 15 rats) received no MNU treatment, but intravesical saline and DMSO once a week for 17 weeks. Group 7 (10 rats) received no treatment and served as normal control. When silibinin and MNU were instilled in the same week, they did not present together in the same intravesical solutions on the same day instead of separation for at least 3 days.

In another cohort study, silibinin was instilled intravesically beginning at week 10 for 8 weeks after 4-dose instillation of MNU with the same grouping, animal numbers, and doses as the first study above (Supplementary Fig. S1). For the treatment, the rats remained anesthetized for 2 hours after catheterization without urination. All animals received medicated water (a combination of trimethoprim sulfamethoxazole, neomycin, and polymyxin B; ref 32). Body weight and diet consumption were recorded weekly throughout the study.

All rats were killed at week 18 and necropsy was performed. The bladders were excised and immediately fixed in 4% paraformaldehyde for at least 24 hours, embedded in paraffin, and submitted for histopathology. Sections were cut transversely through the mid-portion of the bladder and stained with hematoxylin and eosin (H&E).

After desired treatment, the cells and paraffin-embedded sections of samples were studied by terminal deoxynucleotidyl transferase-mediated dUTP nick-end-labeling (TUNEL) assay. Staining was carried out according to the protocol provided by the supplier. Apoptosis was evaluated by counting the positive cells as well as the total number of cells at 10 arbitrarily selected fields at 400× magnification in a blinded manner and quantified as number of apoptotic cells × 100/total number of cells.

Immunohistochemistry (IHC) was conducted using a Dako Autostainer Plus system. Tissues were de-paraffinized, rehydrated, and subjected to 5 minutes of pressure cooking antigen retrieval, 15 minutes of endogenous enzyme block, 60 minutes of primary antibody incubation, and 30 minutes of DakoCytomation EnVision-HRP reagent incubation for mouse and rabbit antibodies, or 15 minutes of biotinylated link coupled with 15 minutes of streptavidin-HRP incubation for goat antibodies. Signals were detected by adding substrate hydrogen peroxide using diaminobenzidine (DAB) as a chromogen followed by hematoxylin counterstaining. Negative control slices were prepared by omitting the primary antibody. Stained (brown) cells were quantified as number of positive cells × 100/total number of cells in 10 random microscopic (400×) fields in each slice.

All statistical analyses were performed using SPSS 15.0. Quantitative data are presented as mean ± SE and the differences among the control and various treatment groups were compared by 1-way ANOVA, followed by Dunnett's t test for separate comparisons. When the comparison involved only 2 groups, Student's t test was used. Differences in the incidence of rat bladder lesions among groups were analyzed by χ² test or Fisher's exact probability test. P < 0.05 was considered statistically significant.

First, we investigated the growth-inhibitory and apoptotic effects of silibinin on bladder cancer cells. As shown in Supplementary Fig. S2A, silibinin inhibited the growth of 5637 cells in a dose- and time-dependent manner. Silibinin at 100, 200, and 400 μM exerted significant inhibitory effects on cell growth (P < 0.05). The effects of silibinin on the cell apoptosis, as measured by flow cytometry, are shown in Supplementary Fig. S2B, with the 48-hour silibinin treatments at 50, 100, and 200 μM resulting in 5.7%, 30.23%, and 47.55% of apoptotic cells, respectively, and the baseline of the control cells being 3.88% (P < 0.05). Correspondingly, the number of TUNEL-positive apoptotic cells increased significantly, following treatment with 100 and 200 μM of silibinin for 48 hours compared with control and 50 μM of silibinin (Supplementary Fig. S2C and S2D).

Mitochondrial membrane potential collapse plays an important role in triggering various apoptotic pathways (33). Using a cationic lipophilic dye JC-1 as a marker of ΔΨ_m, flow cytometric studies revealed that silibinin induced a dose-dependent loss of ΔΨ_m (Fig. 1B). In addition, 80 μM of z-VAD-fmk, a pan caspase inhibitor, showed no protective effects on the dissipation of ΔΨ_m in the cells treated with 200 μM silibinin for 48 hours.

The role of mitochondrial disruption in apoptosis is suggested to be mediated by the releasing of apoptogenic molecules (33). In the present study, we found that the protein levels of mitochondrial cyto c, Omi/HtrA2 and AIF decreased in silibinin-treated cells (Fig. 1C, center). Concomitantly, compared to untreated group, silibinin treatment strongly increased the amount of cyto c, Omi/HtrA2, and AIF in the cytosolic fraction, indicating a release from the mitochondria into the cytosol (Fig. 1C, left). However, no release of Smac/DIABLO was observed. In addition, silibinin had no effects on total levels of cyto c, AIF, Omi/HtrA2, Smac/DIABLO, and 3 individual loading controls (COX IV, GAPDH, and Histone H1 for mitochondrial, cytosolic, and nuclear loading control, respectively, Fig. 1D).

Based on increased apoptosis and release of cyto c into cytosol in silibinin-treated cells, our next aim was to examine the involvement of caspases that are considered to be executioners of apoptosis. As shown in Fig. 2A, silibinin activated caspase-3 and caspase-9, as observed by its increased cysteine protease activity for individual substrate, in a dose-dependent manner. In addition, silibinin increased the cleaved subunits of caspase-3 (17 and 19 kDa), caspase-9 (35 kDa), caspase-8 (10 kDa), and PARP (89 kDa), indicating that silibinin induced the activation of caspase cascade (Fig. 2B). In this study, we also found that silibinin resulted in a decrease in survivin protein level (Fig. 2B). Furthermore, z-VAD-fmk did not completely reverse silibinin-induced apoptosis (Fig. 2C). The specificity and efficacy of z-VAD-fmk activity in 5637 cells was confirmed by caspase-3 activity assay that clearly demonstrated that pretreatment with z-VAD-fmk decreased caspase-3 activity (Fig. 2D). These findings suggested that silibinin-induced apoptosis in 5637 cells was partially caspase dependent.

Because we had observed the release of AIF from mitochondria to cytosol, we next analyzed AIF levels in nuclear protein extracts. We found that 48 hours of exposure to silibinin increased the amount of AIF in the nuclear fractions (Fig. 1C, right). Our results from immunofluorescence staining (Supplementary Fig. S3) also supported the findings. Untreated cells had AIF mainly localized to the cytosol, and the treatment with 200 μM of silibinin for 48 hours resulted in remarkable translocation of AIF to the nucleus. Pretreatment with pan caspase inhibitor was unable to prevent AIF localization to the nucleus. These results supported the contention that silibinin was able to induce caspase-independent apoptosis in the bladder cancer cells.

To extend the observations made in cultured cells and to assess the in vivo efficacy of silibinin, its effects on the growth of 5637 human bladder tumor xenografts were determined in athymic nude mice. All mice survived at the end of the experiment. As shown in Fig. 3A, the growth of 5637 tumor xenografts was inhibited significantly following oral silibinin treatment at the dose levels of 200 and 300 mg/kg. The average tumor masses in the control rats were 2- to 3-fold (P < 0.05) greater than that of 200 and 300 mg/kg of silibinin-treated mice (Fig. 3B). The average body weights and diet consumption of the control and silibinin-treated mice showed no difference throughout the experiment (Fig. 3C and D). Gross and microscopic examination of the liver, lung, spleen, and kidney did not reveal any evidence of change (data not shown). These results suggested the in vivo antitumor efficacy of oral silibinin against human bladder tumor without host toxicity.

As shown in Fig. 4A–D, TUNEL staining showed the number of apoptotic cells increased from 8.5% in the control group to 12.5%, 21% (P < 0.05), and 37.5% (P < 0.05) in 100, 200, and 300 mg/kg of silibinin-treated groups, respectively, demonstrating significant effects of silibinin on apoptosis in vivo. Furthermore, the microscopic examination of stained tumor sections showed strong immunoreactivity for survivin (brown color) in the control group that was decreased in the silibinin-treated groups (Fig. 4I–L). The increased number of cleaved caspase-3–staining cells was also observed in the silibinin-fed group of tumors (Fig. 4E–H). In addition, silibinin treatment at the dose levels of 200 and 300 mg/kg strongly increased the nuclear AIF expression as compared with control and the 100 mg/kg dose of silibinin group (Fig. 4M–P). Taken together, the results clearly showed that silibinin induced both caspase-dependent and -independent apoptosis in vivo.

Because intravesical administration of chemotherapeutic agent is widely used as an adjuvant therapy to prevent recurrence and progression of superficial bladder cancer after TUR, and silibinin was observed to be effective in inhibiting the growth of bladder cancer cells in vitro and in vivo, we next investigated the intravesical efficacy of silibinin in an autochthonous animal model of bladder cancer induced by intravesical instillation of MNU. Approximately 6% (8 of 140) of all animals did not survive all doses in this study. The causes of death were related to the ulceration of the bladder or urosepsis secondary to urethral structure formation and urinary obstruction. Four doses of intravesical MNU to rats resulted in the induction of hyperplasia, papillary dysplasia, atypia, superficial, and muscle invasive bladder carcinoma at the end of the 17-week study (Fig. 5C, Table 1). There were significant differences in histopathologic changes among the silibinin-treated groups and the MNU or MNU + DMSO group (P < 0.05). The incidences of superficial and invasive transitional cell carcinoma were reduced with 200 mg/kg and 400 mg/kg silibinin treatments when compared to MNU or MNU + DMSO group (Table 1). Importantly, there were no invasive lesions in the 400 mg/kg silibinin treatment group. Intravesical silibinin alone did not have any effects on bladder histology (Fig. 5C and D, Table 1). We did not observe any gross sign of toxicity with intravesical silibinin as measured by body weight and diet consumption throughout the study, where there was no significant change in body weight gain (Fig. 5B) and diet intake (Supplementary Fig. S4A) profiles among each group. In addition, in vivo apoptotic effects of intravesical silibinin on MNU-induced bladder cancer were observed by TUNEL staining (Fig. 5D and Supplementary Fig. S4B).

In another cohort study arming to investigate the chemotherapeutic efficacy of intravesical silibinin, approximately 7% (10 of 140) of all animals did not survive all doses in this study. As shown in Table 2, only 400 mg/kg intravesical silibinin treatment beginning at week 10 effectively reduced the incidence of superficial and invasive bladder lesions initiated by MNU when compared to MNU or MNU + DMSO group (P < 0.05). However, there was still one rat (4%) with invasive bladder lesions in this group. Similarly, intravesical silibinin also significantly increased the number of TUNEL-positive apoptotic cells compared with the control group (data not shown).

Taken together, these results showed the chemopreventive and chemotherapeutic effects of intravesical silibinin against MNU-induced bladder carcinogenesis in rats, which may be associated with its proapoptotic effects.

The goals of the present study were to show the in vivo chemopreventive and chemotherapeutic intravesical effects of silibinin on bladder cancer and to elucidate the novel underlying mechanisms of action. Collectively, we have made the following major discoveries. First, silibinin activated the mitochondrial pathways and induced both caspase-dependent and -independent apoptosis in human high-risk bladder carcinoma cells. Silibinin-induced apoptosis involved (a) dissipation of ΔΨ_m; (b) selective releases of cyto c, Omi/HtrA2, and AIF from mitochondria into cytosol; (c) downregulation of survivin; (d) activation of caspases and PARP cleavage, and (e) translocation of AIF to the nucleus. Second, oral administration of silibinin inhibited the growth of human bladder tumor xenografts in athymic nude mice without any side effects. Third, intravesical administration of silibinin effectively inhibited the carcinogenesis and progression of bladder cancer in rats initiated by MNU. Finally, the in vivo anticancer effects of silibinin were shown to link to the in vivo induction of apoptosis and protein expression similar to that seen in vitro. Our studies thus provide a rationale for the development of silibinin as a chemopreventive and chemotherapeutic intravesical agent against bladder cancer in the clinical setting.

It is becoming clear that apoptosis plays a vital role in efficacy of anticancer agents. There are 2 major pathways that initiate apoptosis: one is mediated by death receptors on the cell surface, and the other is mediated by mitochondria (23). The mitochondria have been considered to act as a point of integration for apoptotic signals originating from both of these 2 apoptotic pathways. Among various proapoptotic events and factors, the disruption of ΔΨ_m and release of cyto c are considered critical in mitochondria-mediated apoptosis pathways (34). In this study, silibinin showed a dissipation of ΔΨ_m in bladder cancer cells as evidenced by JC-1 dye staining, which was accompanied by the release of cyto c from mitochondria to cytosol, as evidenced by immunoblotting, and further activation of downstream effectors caspase-9, caspase-3 and resultant PARP cleavage. Activation of caspase-8, a key event in death receptor-mediated apoptosis, also has been shown to lead to the release of cyto c from mitochondria (35). In this study, activation of caspase-8 was also observed. Furthermore, studies are needed to investigate the relationship between caspase-8 activation and the intrinsic mitochondrial pathway in silibinin-treated bladder cancer cells.

Survivin, a member of the inhibitors of apoptosis proteins (IAP) family, has been considered as a molecular signature of unfavorable outcome, a diagnostic biomarker of tumor onset and recurrence, and a validated target for cancer drug discovery (36, 37). Survivin is highly expressed in human bladder carcinomas, and its expression has been correlated with resistance against therapy and abbreviated patient survival (38). Our data showed a strong downregulation of survivin by silibinin. Furthermore, studies are needed to define the underlying molecular mechanisms.

We also found that the inhibitor of caspases, z-VAD-fmk, at a concentration that nearly completely blocked caspase activity, failed to block the silibinin-induced disruption of ΔΨ_m in bladder cancer cells, suggesting caspase-independent pathways of silibinin-induced apoptosis. It is reported that pretreatment with z-VAD-fmk almost completely reversed silibinin-caused cell death in human bladder carcinoma RT4 cells (19). Such disparate findings from our results could be due to the difference in cell type used, implying that there may be fundamental differences of the individual gene expression pattern among various cell lines. Therefore, precautions should be taken when extrapolating results from one to another study.

AIF translocates from mitochondria via the cytosol to the nucleus, where it induces chromatin condensation and DNA fragmentation, and is a hallmark of caspase-independent apoptosis (39). Our data showed that silibinin treatment induced AIF release from mitochondria and translocation to the nucleus, supporting that silibinin is able to induce caspase-independent apoptosis. Thus, the mitochondrial apoptotic pathway involves both cyto c (caspase dependent) and AIF (caspase independent) processes in silibinin-induced apoptotic cell death. In addition, the release of cyto c is usually associated with Smac/DIABLO and Omi/HtrA2 (40), which are also mitochondrial intermembrane space proteins in the apoptotic process. Smac/DIABLO and Omi/HtrA2 have the ability to bind and antagonize the actions of IAPs, which then facilitate the activation of caspases. Cytosolic Omi/HtrA2 also contributes to both caspase-dependent and -independent apoptosis (41). However, in our study we failed to detect Smac/DIABLO release, whereas cyto c and Omi/HtrA2 were translocated to cytosol in silibinin-treated cells. These observations are surprising because several papers have reported that the mitochondrial intermembrane proteins, cyto c, Smac/DIABLO and Omi/HtrA2 are released together with the same or similar kinetic pattern (40,42). Furthermore, studies are needed to identify the relationship between cyto c and Omi/HtrA2 releasing in silibinin-treated cells.

The possibility that bladder cancer cells may be left behind, whether from implantation or incomplete resection, has led many urologists to believe that TUR, though effective, will not be a sufficient treatment alone for all cases of superficial bladder cancer (43). As a consequence, intravesical chemotherapy becomes an important treatment for superficial transitional cell carcinoma. There are 3 fundamental purposes of treatment of superficial bladder cancer with intravesical therapy: (a) eradicating existing lesions, (b) prevention of recurrence, and (c) prevention of tumor progression. The autochthonous animal model of bladder cancer induced by intravesical instillation of MNU utilized in this study has been previously characterized (32). Because DMSO is a urothelium permeability enhancer that was proved to promote the penetration of water-soluble drugs (e.g., cisplatin, pirarubicin, and doxorubicin) and a lipophilic drug paclitaxel across the urothelium in dogs or rats (4), we thus used silibinin dissolved in DMSO as the intravesical agent. In this study, intravesical silibinin beginning either at week 1 or at week 10 effectively inhibited carcinogenesis and progression of bladder cancer in rats by reducing the incidence of superficial and invasive bladder lesions, suggesting its chemopreventive and chemotherapeutic effects against bladder cancer. No toxicity, locally or systemically, was observed in rats receiving intravesical silibinin alone. Intravesical silibinin alone induced 7% and 13% of ordinary hyperplasia instead of atypia hyperplasia or dysplasia in chemopreventive and chemotherapeutic study, respectively. All the chronic cystitis, ulceration of bladder, and urosepsis due to repeatedly intravesical instillation may contribute to this ordinary hyperplasia. As seen in other chemically induced tumor models, the model used in the present study may not completely reflect the clinically observed tumors, yet these are spontaneous, not implanted tumors, and histologically equivalent to the human transitional cell carcinoma seen clinically. Our results showed that intravesical silibinin was both safe and effective in the prevention and therapy of MNU-induced rat bladder cancer.

In summary, our in vitro and in vivo results demonstrated that silibinin induced-apoptosis in human bladder cancer was mediated by the activation of 2 mitochondrial death pathways, namely the cyto c/caspase-dependent and the AIF/caspase-independent pathways involving selective translocation of Omi/HtrA2, and by the downregulation of survivin that is a molecular target of human bladder cancer. Silibinin may prove to be a new form of intravesical chemotherapy in the inhibition of carcinogenesis and progression of bladder cancer, providing a basis for future clinical trials of intravesical silibinin used in patients with bladder cancer.

No potential conflicts of interest were disclosed.

We thank Prof. Erwin G Van Meir, Paula M. Vertino (Emory University School of Medicine, Atlanta, GA) and Jer-Tsong Hsieh (University of Texas Southwestern Medical Center, Dallas, TX) for helpful discussion.

The “13115” Scientific and Technological Innovation Project of Shaanxi Province, China (2008ZDKJ-63)

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.