Food-derived polyphenols inhibit pancreatic cancer growth through mitochondrial cytochrome C release and apoptosis

Abstract

There is increasing evidence that food-derived polyphenols have a beneficial effect for cancers. Our purpose was to determine the effect and mechanism of action of these compounds on pancreatic cancer. We measured effects of quercetin on pancreatic cancer in a nude mouse model. We also investigated the effects of quercetin, rutin, trans-resveratrol and genistein on apoptosis and underlying signaling in pancreatic carcinoma cells in vitro. Quercetin decreased primary tumor growth, increased apoptosis and prevented metastasis in a model of pancreatic cancer. In vitro quercetin and trans-resveratrol, but not rutin, markedly enhanced apoptosis, causing mitochondrial depolarization and cytochrome c release followed by caspase-3 activation. In addition, the effect of a combination of quercetin and trans-resveratrol on mitochondrial cytochrome c release and caspase-3 activity was greater than the expected additive response. The inhibition of mitochondrial permeability transition prevented cytochrome c release, caspase-3 activation and apoptosis caused by polyphenols. Nuclear factor-κB activity was inhibited by quercetin and trans-resveratrol, but not genistein, indicating that this transcription factor is not the only mediator of the polyphenols' effects on apoptosis. The results suggest that food-derived polyphenols inhibit pancreatic cancer growth and prevent metastasis by inducing mitochondrial dysfunction, resulting in cytochrome c release, caspase activation and apoptosis. © 2002 Wiley-Liss, Inc.

AMC, 7-amino-4-methylcoumarin; DiOC6(3), 3,3′-dihexyloxacarbocyanine; DTT, dithiothreitol; ECL, enhanced chemiluminescence; EMSA, electrophoretic mobility shift assay; MG-132, carbobenzoxy-L-leucyl-L-leucyl-L-leucinal (Z-Leu-Leu-Leu CHO); NF-κB, nuclear factor κB; PARP, Poly(ADP-ribose) polymerase; PMSF, phenylmethyl-sulphonyl fluoride; SDS, sodium dodecyl sulfate; SDS-PAGE, SDS polyacrylamide gel electrophoresis; TBS, Tris-buffered saline; Z-DEVD-AMC, Z-Asp-Glu-Val-Asp-AMC; Z-VAD-FMK, Z-Val-Ala-Asp(OMe)-CH2F.

Pancreatic cancer represents the fifth leading cause of cancer deaths in the United States. Cures for this type of cancer are unusual with cancer recurrence as metastatic disease in most cases after removal of the primary tumor by surgery.1, 2

There is increasing evidence from population studies that a diet rich in fruits and vegetables has a beneficial preventative effect for a variety of cancers including pancreatic cancer.3–6 The mechanisms of these beneficial effects on cancer are not known although there is evidence that polyphenolic compounds in these foods may provide some of the benefit. The polyphenols are found in grapes, wine, green tea, soybeans and various fruits. Polyphenols can be divided into 2 groups: the flavonoids and the nonflavonoids.5 Common flavonoids include quercetin, rutin and genistein. Examples of nonflavonoid polyphenolics include the resveratrol family of compounds. Trans-resveratrol has recently received significant attention as a component in red wine and grapes that has antitumor and anti-inflammatory properties.7, 8

In order to determine if the polyphenolic compounds in foods have a beneficial effect on pancreatic cancer, we performed a series of in vivo and in vitro experiments to determine the effects and mechanisms of action of a select group of polyphenolic compounds on pancreatic cancer growth.

To test the differential effects of the polyphenolic compounds on primary vs. metastatic cancer lesions, we chose an orthotopic nude mouse model of metastatic pancreatic cancer utilizing a highly malignant human-derived pancreatic cancer cell line, Mia PACA-2.9 These cancer cells contain mutated p53 and are resistant to apoptosis.10 The model is associated with both growth of the primary tumor in the pancreas and metastases to multiple organs.9

Because the result of our in vivo studies suggested that the beneficial effects of quercetin on tumor growth was due to apoptosis of the cancer cells, we designed a series of in vitro experiments to confirm that the polyphenols cause apoptosis of pancreatic cancer cells as well as determine the intracellular signals involved.

Dysregulation of apoptosis is emerging as a key factor in the mechanism of cancer growth.11, 12 Apoptosis is characterized by morphologic changes including cell shrinkage, chromatin condensation and oligonucleosomal DNA cleavage followed by cell death.13, 14 A critical step in the process of apoptosis is activation of caspases, a specific class of cysteine proteases.11, 15–17

Recent evidence indicates that there are at least 2 distinct pathways that mediate caspase activation and apoptosis.15–22 One involves the ligation of “death receptors” (e.g., TNFα-R1 and Fas), resulting in activation of caspase-8.16–20 This “initiator” caspase activates downstream, or “effector” caspases such as caspase-3. In the second pathway, various forms of cellular stress result in mitochondrial alterations: mitochondrial membrane depolarization and release of cytochrome c. Cytochrome c activates caspase-9, leading to activation of effector caspases.

A major pathway that mediates mitochondrial alterations in apoptosis is the opening of mitochondrial permeability transition pore (MPT11, 23), MPT inhibitor, cyclosporin A, inhibits cytochrome c release in various cell types, which, in addition, may require the phospholipase A₂ inhibitor, aristolochic acid, that prolongs and enhances the inhibitory effect of cyclosporin A.11, 21, 23, 24 Another signaling mechanism that may be important in the regulation of apoptosis is the activation of the nuclear transcription factor, nuclear factor-κB (NF-κB). Activated NF-κB inhibits apoptosis in several cell types.25–28 The effects of NF-κB on apoptosis are thought to be mediated by several gene products regulated by NF-κB including members of a family of genes called inhibitors of apoptosis (IAPs) and members of the Bcl-2 family of genes.27, 28

Our approach for the in vitro experiments in our study involved both demonstrating an effect of the polyphenolic compounds on apoptosis as well as demonstrating the intracellular pathways involved. Our findings revealed that each of the polyphenolic compounds except rutin caused apoptosis that was associated with mitochondrial membrane depolarization and cytochrome c release followed by activation of caspase-3. The MPT inhibitors (cyclosporin A and aristolochic acid) blocked the ability of the polyphenolic compounds to cause cytochrome c release. Cyclosporin A alone blocked cytochrome c release with the flavonoids, quercetin and genistein. In contrast, both cyclosporin A and aristolochic acid were required to block cytochrome c release caused by the nonflavonoid, trans-resveratrol. Furthermore, a flavonoid (quercetin) and nonflavonoid (trans-resveratrol) in combination had a synergistic effect on mitochondrial cytochrome c release and caspase-3 activitation.

The data suggest that cytochrome c release mediates polyphenol-induced apoptosis both in vivo and in vitro. The apoptosis in vivo results in slowing of the growth of the primary tumor and prevention of the development of metastatic lesions.

MATERIAL AND METHODS

Reagents

Rabbit polyclonal antibody against caspase-3 and rabbit polyclonal antibody against poly (ADP-ribose) polymerase (PARP) were from Santa Cruz Biotechnology (Santa Cruz, CA). Poly [d (I-C)] was from Boehringer Mannheim (Indianapolis, IN). Dulbecco's modified Eagle's media, RPMI 1640 and 199 media were from Gibco BRL (Grand Island, NY). Fetal bovine serum (FBS) was from Gibco (UK).32 P-labeled dCTP (3,000 mCi/mmol) and the ECL detection kit were from Pierce (Rockford, IL). The proteosome inhibitor, Z-LLLH (MG-132), was from Peptide International, Inc. (Louisville, KY). Caspase-3 substrate, Ac-Asp-Glu-Val-Asp-AMC and caspase inhibitor, Z-VAD FMK, were from Calbiochem (San Diego, CA). Protein dye was from Bio-Rad Laboratories (Hercules, CA). All other chemicals were from Sigma Chemical Co. (St. Louis, MO).

In vivo experiments

Tumor induction in nude mice was performed as described recently.9 For subcutaneous tumor formation 1×10⁷ Mia PACA-2 tumor cells were subcutaneously injected in the medio-dorsal region of a nude mouse. After 4 weeks a small tumor fragment (∼ 1 mm in diameter) was removed from the subcutaneous tumor and transplanted into the pancreatic tail of mice of the 2 study groups. Ten days after tumor transplantation, treatment with or without quercetin was initiated. The treated animals received daily intraperitoneal injections of 1.3 mg quercetin dissolved in DMSO; the control group received DMSO only by intraperitoneal injection. The animals were sacrificed once clinical tumor signs including severe cachexia, ascites with abdominal distension or heavy tumor burden (larger than 1.5 cm) became apparent.

Tumor volume was calculated as described previously.9) as 0.5 × length × width × depth. Metastatic tumor spread was determined macroscopically at autopsy in all thoracic, abdominal, retroperitoneal and pelvic organs. All macroscopic suspicious lesions were further confirmed as tumor dissemination by microscopic analysis. Each value in the metastatic score represented a different organ of metastatic tumor spread.

Cell lines

Human pancreatic carcinoma cell line Mia PACA-229 and rat pancreatic carcinoma BSp73AS30 were cultured in Dulbecco's Modified Eagle's Medium supplemented with 10% heat inactivated FBS, penicillin G (100 U/ml) and streptomycin (100 mg/ml) in a humidified atmosphere containing 5% (v/v) CO₂. When cells were 90% confluent they were detached, washed with alternating centrifugation and resuspension, plated and incubated in the same media without FBS and with the indicated concentrations of polyphenolic compounds or vehicle for up to 96 hr.

Apoptosis measurement in human xenograft tumors

Apoptosis was measured on paraffin-embedded tissue sections using the TUNEL assay.31–33 Tissue sections (3 μm) were deparaffinized and rehydrated through a graded series of ethanol and redistilled water. Tissue sections were refixed in 4% paraformaldehyde for 15 min at room temperature and then incubated with proteinase K (20 μg/ml in 10 mM Tris/HCl, pH 7.4–8.0) for 15 min at 37°C. DNA breaks were then labeled with terminal deoxytransferase (TdT) and biotinylated deoxyUTP. Staining without TdT enzyme or the biotinylated substrate were used as negative controls. For positive controls slides were treated with DNase I.

DNA isolation

DNA was isolated as described previously.31 Briefly, pancreatic cancer cells growing on plates were removed by treatment with trypsin, collected by centrifugation and lysed by resuspension in a buffer containing 10 mM Tris/HCl (pH 8.0) 10 mM NaCl, 10 mM EDTA, 300 μg/ml proteinase K and 1% SDS. Cell lysates were incubated overnight at 45°C; and DNA was purified by phenol/chloroform extraction (1:1 vol/vol), precipitated overnight at 20°C with 0.3 M sodium acetate and collected by centrifugation at 15,000 g for 15 min at 4°C. The pellet containing RNA and DNA was resuspended in TE buffer [10 mM Tris/HCl (pH 8.0), 1.0 mM EDTA] and treated subsequently with RNase (200 μg/ml) for 2 hr at room temperature, followed by an incubation overnight with proteinase K (200 μg/ml) at 45°C. Finally, the mixture was re-extracted with phenol/chloroform and chloroform, precipitated with ethanol and resuspended in TE buffer. DNA fragments were separated electrophoretically on 1.8% agarose gel containing 0.5 μg/ml ethidium bromide in 0.5 × TBE buffer (TBE: 89 mM Tris base, 89 mM boric acid and 2 mM EDTA).

Annexin-V staining

Approximately 1 × 10⁶ cells (as determined with a hemocytometer) were analyzed for annexin-V binding using an Annexin-V-FLUOS Staining Kit (Boehringer Mannheim, Germany). Briefly cells were washed twice with PBS and incubated for 10 min at room temperature with fluorescein isothiocyanate (FITC)-conjugated, annexin-V reagent (20 μg/ml) and propidium iodide (50 μg/ml). Cells were analyzed on a FACScan flow cytometer (Becton Dickinson Immunocytometry System, San Jose, CA) equipped with a 15 nW air-cooled 488 nm argon-ion laser. Annexin-V positive and propidium iodide negative cells were considered as apoptotic.

Caspase-3 activity

Cells were collected, washed with ice-cold PBS and resuspended in the lysis buffer containing 0.5% Nonidet P-40 or manufactured by the name IGEPAL CA-630, 0.5 mM EDTA, 150 mM NaCl and 50 mM Tris at pH 7.5. Cell lysates were placed for 30 min on a rotator at 4°C and then centrifuged for 15 min at 15,000g. Cytosolic protein extracts (supernatants) were collected, protein concentrations were determined and the extracts were aliquoted and stored at −80°C. Enzyme assays were carried out at 37°C in a buffer containing 25 mM HEPES (pH 7.5), 10% sucrose, 0.1% CHAPS and 10 mM DTT with 800g cytosolic protein and 20 μM of specific fluorogenic substrate. For caspase-3, the substrate was Z-Asp-Glu-Val-Asp-AMC (z-DEVD). Cleavage of the caspase substrate releases AMC (7-amino-4-methylcoumarin), which emits a fluorescent signal with excitation at 380 nm and emission at 440 nm. The reaction was started by addition of caspase-3 substrate, the readings were taken at 0, 60, 90 and 120 min. Fluorescence was calibrated using a standard curve for AMC. The data are expressed as mol AMC/mg protein/min.

Western blot analysis

To extract proteins, cells were washed twice with PBS and lysed by incubating for 20 min at 4°C in lysis buffer containing 0.15 M NaCl, 50 mM Tris (pH 7.2), 1% deoxycholic acid (wt/vol), 1% Triton X-100 (wt/vol), 0.1% SDS (wt/vol) and 1 mM PMSF, as well as 5 μg/ml each of protease inhibitors, pepstatin, leupeptin, chymostatin, antipain and aprotinin. Then the cell lysates were centrifuged for 20 min at 15,000g at 4°C. The supernatants were separated by 4–20% SDS-PAGE for 2 hr at 120 V using precast Tris-glycine gels and a Mini-Cell gel apparatus (Novex, San Diego, CA). Separated proteins were electrophoretically transferred to a nitrocellulose membrane for 2 hr at 30 V using a Novex Blot Module. Nonspecific binding was blocked by 1 hr incubation of nitrocellulose membranes in 5% (wt/vol) nonfat dry milk in Tris-buffered saline (TBS; pH 7.5). Blots were then incubated overnight at 4°C with anti- caspase-3 antibody (1:3,000) in an antibody buffer containing 1% (wt/vol) non-fat dry milk in TTBS (0.05% vol/vol Tween-20 in TBS), washed 3 times with TTBS and finally incubated for 1 hr with a peroxidase-labeled secondary antibody in the antibody buffer. Blots were developed for visualization using ECL detection kit. To test for equal protein loading, the blots were stripped and re-probed with an antibody against tubulin. When processing of a protein (e.g., caspase or PARP) was measured, a decrease in unprocessed full-length form was measured concomitantly with the increase in the cleaved, active form (Figs. 1,2,4).

Preparation of mitochondrial and cytosol fractions

Cells were washed twice with ice-cold PBS, pH 7.2 and resuspended in extraction buffer (∼ 500 μl) containing 20 mM HEPES-KOH (pH 7.0), 10 mM KCl, 1 mM NaEGTA, 2 mM MgCl₂, 1 mM EDTA, 1 mM DTT, 250 mM sucrose, 1 mM PMSF and protease inhibitors cocktail listed in the Western blot analysis section. Lysate was incubated 30 min on ice and then homogenized using a glass dounce (80 strokes). Nuclei were removed by centrifugation at 1,000g for 10 min at 4°C. Supernatant was additionally centrifuged for 1 hr at 100,000 x g and the resulting supernatant (cytosolic fraction) and pellet (mitochondrial fraction) were collected separately and used for Western blotting.

Mitochondrial membrane potential

Mitochondrial membrane potential was determined as described34 by measuring the retention of the dye 3,3′-dihexyloxacarbocyanine (DiOC₆³. Cells were loaded with 1 μM DiOC₆³ during the last 30 min of treatment with a polyphenolic compound (or vehicle). The cells were then collected and pelleted by centrifugation. The supernatant was removed and the pellet was washed twice with PBS by alternate centrifugation and resuspension. The pellet was then lysed by addition of 1 ml of H₂O and homogenized. The concentration of DiOC₆³ was read on a Perkin-Elmer LS-5 fluorescence spectrometer at 488 nm excitation and 500 nm emission.

Preparation of nuclear extracts

Nuclear protein extracts were prepared essentially as described before.35–37 Briefly, cells were lysed on ice in the hypotonic buffer A supplemented with 1mM phenylmethylsulfonyl fluoride (PMSF) and 1 mM dithiothreitol (DTT) and with the protease inhibitor cocktail containing 5 μg/ml each of pepstatin, leupeptin, chymostatin, antipain and aprotinin by 20 strokes in a glass Dounce homogenizer. The homogenate was incubated on ice for 15–20 min and then 0.3% (vol/vol) Nonidet P-40 was added and the samples were briefly vortexed and incubated on ice for an additional 2 min. Crude nuclear pellet was collected by centrifugation. The supernatant (cytosolic protein) was saved for Western blot analysis. The nuclear pellet was resuspended in the high-salt buffer C38–40 containing 20 mM HEPES (pH 7.6), 25% (vol/vol) glycerol, 0.42 M NaCl, 1.5 mM MgCl₂, 0.2 mM EDTA, 20 mM B-glycerophosphate, 10 mM Na₂MoO₄, 50 mM Na₃VO₄, 1mM DTT, 1 mM PMSF and the protease inhibitor cocktail described above. After rotation at 4°C for up to 1 hr, the nuclear membranes were pelleted by microcentrifugation for 10 min and the clear supernatants (nuclear extracts) were aliquoted and stored at −80°C.

Electrophoretic mobility shift assay

Aliquots of nuclear extracts with equal amount of protein (2–10 μg) were mixed in 20 μl reactions with a buffer containing 10 mM HEPES (pH 7.6), 50 mM KCl, 0.1 mM EDTA, 1 mM DTT, 10% (vol/vol) glycerol and 3 μg poly[d (I-C)]. After aliquots were equilibrated on ice for 5 min, binding reactions were started by addition of 20–60,000 counts/min (20 fM) of ³²P-labeled DNA probes and allowed to proceed for 25–30 min at room temperature or up to 1 hr on ice. The oligonucleotide probe 5′-GCAGAGGGGACTTTCCGAGA containing a κB binding motif (underlined) was annealed to the complementary oligonucleotide with a 5′-G overhang and end-labeled using Klenow DNA polymerase I. Samples were electrophoresed at room temperature in 0.5 × TBE buffer (1 × TBE 89 mM Tris base, 89 mM boric acid and 2 mM EDTA) on nondenaturing 4.5% polyacrylamide gel at 200 V. Gels were dried and directly analyzed in the PhosphorImager (Molecular Dynamics, Sunnyvale, CA).

Statistical analysis

Statistical analysis of data was done using unpaired Student's t-test.

RESULTS

As our first step in determining the effect of the polyphenolic compounds on pancreatic cancer growth, we tested the effect of quercetin in a nude mouse model of pancreatic cancer using the highly malignant pancreatic cancer cell line- Mia PACA-2. As indicated in Table I, the quercetin treatment had multiple effects on the in vivo growth of the tumor. A major finding was that the quercetin treatment prevented metastatic cancer lesions. That is, in the control animals the mean number of organs with metastatic lesions was 4.4 and in the quercetin-treated animals only 0.6.

Table I. Effect of Treatment With Quercetin on Tumor Progression in vivo1

Parameters of cancer development	Control	Quercetin
1 Survival in control animals was measured as the number of days after transplantation until the animal died or appeared severely ill. Survival in the quercetin-treated animals was measured as the number of days after transplantation until the animal appeared ill from abdominal distension. The distension was due to dilation of the small and large bowels. The values represent means ± SE, n = 8.–*p < 0.05 compared to untreated animals.
Survival, days	66.6 ± 3.2	75 ± 2.5*
Tumor volume, cm^3	5.76 ± 0.92	1.56 ± 0.27*
Number of metastatic sites	4.4 ± 0.9	0.62 ± 0.26*
Percentage of apoptosis	3.3 ± 0.6	7.1 ± 1.0*

In addition to its effect on metastatic lesions, the data in Table I indicates that the quercetin treatment significantly decreased the growth of the primary tumor. Of note, we found similar inhibitory effects of genistein on the development of metastatic lesions and the growth of the primary tumor in the same model of pancreatic cancer.9 In order to provide information on the mechanism of the suppressive effects of quercetin on the growth of the pancreatic cancer, we measured apoptosis in the primary tumors using the TUNEL assay. As indicated in Table I, there was a significant increase in the percentage of cells undergoing apoptosis in the quercetin-treated animal compared to the control animals. In contrast to the effect of the quercetin treatment on apoptosis in the tumor tissue, there was no increase in apoptosis detected by the TUNEL assay in normal tissues (i.e., liver) (data not shown). Such a result indicates that quercetin's effects on apoptosis were specific for the tumor tissue.

To confirm the findings of apoptosis in vivo and to initiate the series of studies on the mechanism of apoptosis, we first determined the effects of quercetin, rutin and trans-resveratrol on other measures of apoptosis (oligonucleotide DNA fragmentation, annexin staining and PARP proteolysis) in Mia PACA-2 cells and BSp73AS cells in culture. BSp73AS cells are derived from a rat pancreatic carcinoma. Both Mia PACA-2 and BSp73AS cells have mutated p53 and express K-ras.

As indicated in Figure 1a,b, quercetin and trans-resveratrol but not rutin caused an increase in oligonucleosomal DNA fragmentation, a unique characteristic of apoptosis, as well as increased Annexin staining. Annexin staining is a measure of externalization of phosphatidylserine to the outer plasma membrane leaflet representing another unique characteristic of apoptosis. The dose-response evaluation in Figure 1b indicates that quercetin is more potent in causing apoptosis than trans-resveratrol. Finally, the data in Figure 1c illustrates that a protein target of caspases-3 activation, PARP, was cleaved in cell lines treated with quercetin and trans-resveratrol but not rutin. The cleavage did not occur in the presence of the specific caspase inhibitor, Z-VAD. These findings of apoptosis with quercetin and trans-resveratrol occurred independent of the presence of serum in the incubation media (data not shown).

Figure 1. Polyphenols stimulate apoptosis in pancreatic carcinoma cells. (a) BSp73AS cells were cultured for 6 hr in the presence or absence of 100 μM of each rutin (RN), quercetin (Q) or trans-resveratrol (RS). DNA was extracted from cell samples and DNA fragments were separated with gel-electrophoresis. The experiment was repeated twice with similar results. (b) Mia PACA-2 cells were cultured for 72 hr in the presence of indicated concentrations of rutin, quercetin or trans-resveratrol. Samples were analyzed by annexin staining. The values represent means ± SE (n = 3). *p < 0.05 compared to control cells. (c) BSp73AS were cultured for 6 hr and Mia PACA-2 cells for 24 hr in the presence or absence of 100 μM of each rutin, quercetin or trans-resveratrol and with or without 50 μM of Z-VAD FMK(Z-VAD). Cell lysates were obtained and analyzed using Western blot analysis with antibody against PARP. The experiment was repeated 3 times with similar results.

The results in Figure 1 confirm that quercetin and trans-resveratrol activate apoptosis in pancreatic cancer cells and support the contention that the beneficial effect of quercetin in vivo is due to their ability to cause apoptosis.

The remaining studies in this report were designed to reveal the intracellular mechanisms that mediate the apoptosis caused by the polyphenolic compounds. The results in Figures 2 and 3 are from experiments designed to determine the effects of the compounds on caspase-3 activity. First, both quercetin and trans-resveratrol convert caspase -3 from its inactive form (a 32 kDa doublet) to its active form (17 kDa) as illustrated by a decrease in the inactive form and an increase in the active form using Western blot analysis and an antibody that recognizes both forms (figs. 2a, 3a). The results show a dose dependency with effects occurring with as little as 20 μM for both compounds. Using a specific fluorogenic assay for caspase-3, we confirmed that both quercetin and trans-resveratrol but not rutin cause caspase-3 activation (Figs. 2b,c, 3b). The effects on caspase -3 activity were dose (Figs. 2a,c, 3a) and time (Figs. 2b, 3b) dependent. Of note, the time course results show that effects of quercetin and trans-resveratrol on apoptosis were more rapid in BSp73AS compared to Mia PACA-2 cells. The reasons for this difference are unknown.

Figure 2. Polyphenols activate caspase-3 in BSp73AS cells in a time-and dose-dependent manner. (a) Cells were cultured for 24 hr with indicated concentrations of trans-resveratrol. Cell lysates were prepared and 50 μg protein aliquots were loaded per lane and blotted with antibody against caspase-3. The experiment was repeated 3 times with similar results. (b) Cells were cultured for indicated time in the presence of 100 μM of each quercetin (Q), trans-resveratrol (RS), rutin (RN) or control (cont). Caspase-3 activity was measured in cell lysates with a fluorogenic assay using DEVD-AMC as a substrate. The results normalized to the DEVDase activity in untreated cells. The values represent means ± SE (n = 3). *p < 0.05 compare to control cells. (c) Cells were cultured for 6 hr in the presence of indicated concentrations of quercetin. Caspase-3 activity was measured in cell lysates with a fluorogenic assay using DEVD-AMC as a substrate. The results normalized to the DEVDase activity in untreated cells. The values represent means ± SE (n =;3). *p < 0.05 compare to control cells.

Figure 3. Polyphenols activate caspase-3 in Mia PACA-2 cells in a time-and dose-dependent manner. (a) Mia PACA-2 cells were cultured for 72 hr in the presence or absence of indicated concentrations of quercetin. Cell lysates were prepared and 50 μg protein were loaded per lane and blotted with antibody against caspase-3. Blots were then stripped and re-probed with an antibody against tubulin to confirm equal protein loading. The experiment was repeated 3 times with similar results. (b) Mia PACA-2 cells were cultured for indicated time in the presence of 100 μM of each quercetin (Q), trans-resveratrol (RS), rutin (RN) or control (cont). Caspase-3 activity was measured in cell lysates with a fluorogenic assay using DEVD-AMC as a substrate. The results normalized to the DEVDase activity in untreated cells. The values represent means ± SE (n =;3). *p < 0.05 compare to control cells.

The results illustrated in Figures 4 though 6 demonstrate the effects of the polyphenols on mitochondrial cytochrome c release, mitochondrial polarity and the role of these effects on apoptosis. For the experiment in Figure 4, we used Western blot analysis and an antibody specific for cytochrome c to probe both cytosolic and mitochondrial fractions to determine if there was a transfer of cytochrome c from the mitochondria to the cytosol. Quercetin, trans-resveratrol and genistein, but not rutin, caused increases in cytosolic cytochrome c and decreases in mitochondrial cytochrome c. Of note, we showed previously that genistein stimulates both apoptosis and caspase-3 activation in Mia PACA-2 cells.

Figure 4. Polyphenols stimulate mitochondrial release of cytochrome c in pancreatic carcinoma cells. BSp73 AS cells (a) were cultured for 6 hr and Mia PACA-2 cells (b) for 24 hr in the absence or presence of 100 μM of each rutin (RN), trans-resveratrol (RS), genistein (GN) or quercetin (Q). Cytosolic and mitochondria extracts were prepared and subjected to SDS-PAGE followed by protein transfer. Immunoblot was performed with antibody against cytochrome c. The experiment was repeated 2 times with similar results.

The results in Figure 5 show that quercetin, trans-resveratrol and genistein, but not rutin, cause dissipation of the mitochondrial membrane potential using a dye, DiOC₆³, that is taken up in cells as a function of mitochondrial membrane potential. Because one mechanism of mitochondrial cytochrome c release involves opening of the mitochondrial membrane permeability transition (MPT), which is associated with dissipation of the mitochondrial membrane potential, the results in Figure 5 suggest that the mechanism of action of the polyphenolic compounds on cytochrome c release and apoptosis is through their effect on the MPT.

Figure 5. Polyphenols induce depolarization of mitochondrial membrane potential in pancreatic carcinoma cells. (a) BSp73AS cells were cultured for 6 hr in the absence (cont) or presence of the indicated concentrations of quercetin (Q) and trans-resveratrol (RS) or during the last 30 min of treatment DiC₆³ was added. An aliquot of the cells was used for determination of retained DiOC₆³ fluorescence retained by the cells. The values represent means ± SE (n = 3). *p < 0.05 compared to control cells. (b) Mia PACA-2 cells were cultured for 24 hr in the presence or absence of 100 μM of each rutin (RN), trans-resveratrol (RS), genistein (GN) or quercetin (Q) or control (cont). During the last 30 min of treatment DiOC₆³ was added. An aliquot of the cells was used for determination of DiOC₆³ fluorescence retained by the cells. Values represent means ± SE, (n = 3). *p < 0.05 compare to control cells.

The results of the experiments in Figure 6 demonstrate the effects of a caspase inhibitor, Z-VAD and agents that act on the MTP, cyclosporin A and aristolochic acid, on cytochrome c release, caspase-3 activity and apoptosis in cancer cells treated with the polyphenolic compounds. Cyclosporin A inhibits MPT channels by interacting with one of the key subunits of the MTP, cyclophilin.11, 23, 24 Cyclosporin A by itself or in combination with aristolochic acid blocks cytochrome c release on several cell types.11, 23, 24 As illustrated in Figure 6a, Z-VAD inhibited the release of cytochrome c into the cytoplasm in control (untreated) cells. In contrast, Z-VAD had no effect on cytosolic cytochrome c release caused by quercetin or trans-resveratrol. The lack of effect of Z-VAD on cytochrome c release caused by these agents indicates that their action is directly on the mitochondria and not through the pathway involving caspase-8 and Bid. The cytochrome c release caused by quercetin was inhibited by cyclosporin A alone, whereas the cytochrome c release caused by trans-resveratrol required both cyclosporin A and aristolochic acid for inhibition.

Figure 6. Inhibition cytochrome c release prevents caspase-3 activity and apoptosis induced by polyphenols. (a) Mia PACA-2 cells were cultured for 24 hr in the presence or absence of quercetin (100 μM), trans-resveratrol (100 μM), Z-VAD FMK (50 μM), cyclosporin A (5 μM) and aristolochic acid (50 μM). Cytosolic extracts were prepared and subjected to SDS-PAGE followed by protein transfer. Immunoblot was performed with antibody against cytochrome c. Blots were then stripped and re-probed with an antibody against tubulin to confirm equal protein loading. The experiment was repeated 2 times with similar results. (b) Mia PACA-2 cells were cultured for 24 hr in the presence or absence of quercetin (Q, 100 μM), resveratrol (RS, 100 μM), genistein (GN, 100 μM), cyclosporin A (5 μM) and aristolochic acid (50 μM). Caspase-3 activity in cell lysates were evaluated with a fluorogenic assay using DEVD-AMC as a substrate. The results were normalized to the DEVDase activity in cells not treated with polyphenols. The values represent means ± SE (n = 3). *p < 0.05 compare to cells treated with the polyphenol alone. (c) Mia PACA-2 cells were cultured for 24 hr in the presence or absence of quercetin (Q, 100 μM), trans-resveratrol (RS, 100 μM), cyclosporin A (5 μM) and aristolochic acid (50 μM). Samples were analyzed by annexin staining. The values represent means ± SE (n = 3). *p < 0.05 compared to untreated cells.

The results in Figure 6b demonstrate the effects of the cyclosporin A and aristolochic acid on caspase-3 enzyme activity. The results indicate that cyclosporin A alone inhibited caspase-3 activity in quercetin and genistein treated cells, whereas both cyclosporin A and aristolochic acid were required to inhibit caspase-3 activity in trans-resveratrol- treated cells. Finally, the measurements in Figure 6c demonstrate that cyclosporin A alone inhibited apoptosis in quercetin-treated cells; whereas both cyclosporin A and aristolochic acid were required to inhibit apoptosis caused by trans-resveratrol. The combined results from Figures 6a–c indicate that the effect of the polyphenolic compounds to cause apoptosis results from their direct effects on the cancer cell mitochondrial MTP to cause cytochrome c release. The differential effects of the MTP inhibitors on cytochrome c release with the flavonoids (quercetin and genistein) vs. the nonflavonoids (trans-resveratrol) suggested that the flavonoids and nonflavonoids interact with the mitochondrial MTP by two separate pathways. In order to provide further evidence for 2 pathways, we determined the effects of combinations of quercetin and trans-resveratrol on cytochrome c release and caspase-3 activity. As illustrated in Figure 7, the combinations resulted in responses of cytochrome c release from mitochondria and caspase-3 activity that were statistically significantly greater than the additive responses. That is, with the concentrations used trans-resveratol had only a slight effect on measures of cytochrome c release and caspase-3 activity. However, the same concentration caused a substantial increase in cytochrome-c release caused by quercetin. Furthermore, this concentration of trans-resveratol more than doubled the effect of quercetin on caspase-3 activity. The synergistic responses with 2 agents support the contention that the flavonoids and nonflavonoids act on the MTP by distinct pathways. Such results also suggest that the pathways are interactive.

Figure 7. Effect of a combination of trans-resveratrol and quercetin on cytochrome-c release and caspase-3 activity. (a) Mia-PACA-2 cells were cultured for 24 hr in the presence or absence of trans-resveratrol (25μM), quercetin (25μM), or the combination of trans-resveratrol (25μM) and quercetin (25μM). Cytosolic extracts were prepared and subject to SDS.PAGE followed by protein transfer. Immunoblot was performed with an antibody against cytochrome c. Blots were then stripped and re-probed with an antibody against tubulin to confirm equal protein loading. (b) Mia-PACA-2 cells were cultivated for 24 hr in the presence or absence of trans-resveratrol (25μM), quercetin (25μM) or the combination of trans-resveratrol (25μM) and quercetin (25μM). Caspase-3 activity was measured in cell lysates with a fluorgenic assay using DEVD-AMC as a substrate. The results were normalized to the DEVDase activity in untreated cells. The values for trans-resveratrol, quercetin and the combination of trans-resveratrol and quercetin represent the means ± SE (n = 3) with the values for the controls subtracted. The dashed line over the bar for the values for trans-resveratrol plus quercetin represents the “predicted” additive values for the response to both agents. The recorded values were statistically significantly greater than the “predicted” additive values (p < 0.05).

The experiments in figure 8 were designed to determine the role of activated NF-κB in the regulation of apoptosis caused by the polyphenolic compounds. As indicated in Figure 8a, NF-κB is constitutively active in both cancer cell lines. Figure 8a,b demonstrates that quercetin and trans-resveratrol inhibit NF-κB activation in both pancreatic cancer cell lines; rutin activates NF-κB in BSp73AS cells but not Mia PACA-2 cells; and genistein has no effect on NF-κB in the Mia PACA-2 cells. The proteosome inhibitor, MG-132,38, 39 blocks NF-κB activation in both cell lines. The results in Figure 8c show that MG causes a small increase in caspase-3 activity that adds to the caspase-3 activity caused by quercetin. Furthermore, complete inhibition of NF-κB activation by MG does not increase apoptosis rates to the same extent as trans-resveratrol which only partially inhibits NF-κB activation. Finally, the results show that genistein causes significant apoptosis in the absence of an effect on NF-κB activation. Thus the results in Figure 8 indicate that inhibition of NF-κB is not the only mechanism by which the polyphenols stimulate apoptosis.

Figure 8. Effect of polyphenols and the proteosome inhibitor, MG-132, on NF-κB activity, activation of caspase-3 and apoptosis. (a) BSp73AS cells were cultured for 6 hr and Mia PACA-2 for 24 hr in the absence (Cont) or presence of polyphenols: rutin (RN), quercetin (Q), trans-resveratrol (RS) or genistein (GN), each at 100 μM and 20 μM proteosome inhibitor MG-132 (MG). Nuclear proteins were isolated and analyzed for NF-κB DNA binding activity with EMSA. Positions of specific NF-κB−DNA complexes and the free probe are indicated by single and double arrowhead, respectively. The experiment was repeated twice with similar results. (b) Relative NF-κB activities in cells treated with polyphenols or MG-132. The values represent densitometric intensities of NF-κB band quantified with PhosphorImager, relative to cells non-treated with polyphenols or NF-κB inhibitors. Values represent means ± SE. (n = 3). *p < 0.05 as compared to control cells. (c) Caspase-3 activity and annexin staining were measured. The values represent means *SE (n = 3). The results for caspase 3 were normalized to the DEVDase activity in untreated cells. *p < 0.05 as compared to control cells.

DISCUSSION

The findings in the our report demonstrate that the polyphenolic food-derived compound, quercetin, has a dramatic effect on the growth and metastasis of pancreatic cancer in a nude mouse model. The quercetin treatment markedly decreased the growth rate of the primary tumor; moreover, it almost completely prevented the development of metastatic lesions. We recently reported similar effects of genistein in the nude mouse model of pancreatic cancer.9

The in vivo studies suggested and the in vitro studies confirmed that the beneficial effects of quercetin and the other polyphenolic compounds were due to their ability to cause apoptosis. Of note, we found that apoptosis occurred in the tumor tissue in vivo and not normal tissues suggesting that actions of the polyphenolic compounds are specific for cancer tissue.

The combined results of the in vitro studies indicate that 3 food-derived polyphenolic compounds (quercetin, trans-resveratrol and genistein) cause apoptosis in pancreatic cancer cells by mechanisms involving their interaction with the cancer cell mitochondrial membrane permeability transition (MPT) complex resulting in cytochrome c release from the mitochondria into the cytoplasm. Although we have not determined the exact details of this interaction, our findings provide several important insights about the interaction. First, the results indicate that these three compounds cause apoptosis of the cancer cells through their effects on the MTP. That is, blockade of the cytochrome c release by MTP inhibitors prevents their effects on both caspase-3 activation and apoptosis. This is a particularly important finding since these compounds have been found to affect other intracellular pathways that could regulate the growth of tumor cells.7, 8, 40–44 In addition, selected polyphenolic compounds have been reported to cause apoptosis in cancer cell lines.45–56 Our findings indicate that at least with respect to apoptosis the effects of the polyphenolic compounds on the mitochondrial MTP supercede any other effects that the agents may have.

A second characteristic of the interaction is that the polyphenolic compounds do not interact with the mitochondria through an activated caspase pathway (i.e., caspase-8) because the broad spectrum caspase inhibitor, Z-VAD, did not alter the mitochonrial cytochrome c release caused by the polyphenolic compounds.

A third important characteristic of the interaction relates to the finding that the effects of the agents on the mitochondria are not due to entirely to effect that they may have on NF-κB activation. Other reports have shown that the polyphenolic compounds, quercetin and trans-resveratrol, inhibit NF-κB activation and increase apoptosis in in vitro studies of cancer cells.55, 56 Our results confirm the findings that these agents can inhibit NF-κB activation and that inhibition of NF-κB activity does increase apoptosis. However, although quercetin and trans-resveratrol inhibited NF-κB activation, genistein produced no effect on NF-κB binding activity. This suggests that NF-κB inhibition is not the only mechanism by which the polyphenolic compounds stimulate apoptosis.

Of further interest, we found that cyclosporin A alone prevented cell death in quercetin and genistein treated cancer cells while both cyclosporin A and aristolochic acid were needed to prevent cytochrome c release and cell death caused by trans-resveratrol. Such results suggest that there are 2 distinct pathways by which polyphenolic compounds cause mitochondrial dysfunction in pancreatic cancer cells. To further explore the possibility that the polyphenolic compounds act by distinct pathways, we tested the effects of combinations of quercetin and trans-resveratrol on cytochrome c release and caspase activation. The results show synergistic effects of quercetin and trans-resveratrol on these measures. Since such the results not only further support the contention that these agents act by separate pathways, they demonstrate interactions between the pathways that result in synergistic actions. This finding has important implications not only related to the mechanism of cell death in cancer cells but also indicates an important beneficial role for combinations of polyphenolic compounds in the diet for cancer prevention and control.

In sum, our findings demonstrate that flavonoid and nonflavonoid polyphenolic compounds found in food inhibit pancreatic cancer growth both in vivo and in vitro by interacting with the mitochondrial MTP of the cancer cells. The interactions lead to mitochondrial depolarization and cytochrome c release. These mitochondrial effects of the polyphenolic compound account, at least in part, for their effects on apoptosis. Of particular interest, combination of a flavonoid and a nonflavonoid compound have synergistic effects on the mitochondria and apoptosis. These synergistic effects suggest that the flavonoids and nonflavonoids act on the cancer cell mitochondria by distinct and interacting pathways. Such results also have important implications for combinations of polyphenolic compounds for cancer prevention.