Fisetin inhibits cellular proliferation and induces

MOLECULAR CARCINOGENESIS

Fisetin Inhibits Cellular Proliferation and Induces Mitochondria-Dependent Apoptosis in Human Gastric Cancer Cells Akash Sabarwal,1 Rajesh Agarwal,2 and Rana P. Singh1,3* 1

School of Life Sciences, Central University of Gujarat, Gandhinagar, Gujarat, India Department of Pharmaceutical Sciences, School of Pharmacy, University of Colorado Denver, Aurora, Colorado 3 Cancer Biology Laboratory, School of Life Sciences, Jawaharlal Nehru University, New Delhi, India 2

The anticancer effects of fisetin, a dietary agent, are largely unknown against human gastric cancer. Herein, we investigated the mechanisms of fisetin-induced inhibition of growth and survival of human gastric carcinoma AGS and SNU-1 cells. Fisetin (25–100 mM) caused significant decrease in the levels of G1 phase cyclins and CDKs, and increased the levels of p53 and its S15 phosphorylation in gastric cancer cells. We also observed that growth suppression and death of non-neoplastic human intestinal FHs74int cells were minimally affected by fisetin. Fisetin strongly increased apoptotic cells and showed mitochondrial membrane depolarization in gastric cancer cells. DNA damage was observed as early as 3 h after fisetin treatment which was accompanied with gamma-H2A.X(S139) phosphorylation and cleavage of PARP. Fisetininduced apoptosis was observed to be independent of p53. DCFDA and MitoSOX analyses showed an increase in mitochondrial ROS generation in time- and dose-dependent fashion. It also increased cellular nitrite and superoxide generation. Pre-treatment with N-acetyl cysteine (NAC) inhibited ROS generation and also caused protection from fisetininduced DNA damage. The formation of comets were observed in only fisetin treated cells which was blocked by NAC pretreatment. Further investigation of the source of ROS, using mitochondrial respiratory chain (MRC) complex inhibitors, suggested that fisetin caused ROS generation specifically through complex I. Collectively, these results for the first time demonstrated that fisetin possesses anticancer potential through ROS production most likely via MRC complex I leading to apoptosis in human gastric carcinoma cells. © 2016 Wiley Periodicals, Inc. Key words: fisetin; gastric cancer; apoptosis; mitochondrial respiratory chain complex I; ROS

INTRODUCTION Gastric cancer is one of the major causes of mortality in both developed and developing countries, with high recurrence rate and poor survival. Approximately, 1 million new gastric cancer cases and 0.7 million deaths were estimated to have occurred in 2012 worldwide, accounting for 8% of the total cases and 10% of total deaths [1]. Recent data suggest that there is a substantial decrease in cases of gastric cancer in most parts of the world [2], in part due to decreased reliance on salted and preserved food and increased availability of fresh fruits and vegetables stored under refrigeration systems. Still, gastric cancer remains the third leading cause of cancer death. In India, 63 000 new cases and 59 000 deaths from gastric cancer were estimated to have occurred [1]. While much has been learned about the genetic and biochemical mechanisms underlying gastric cancer, few novel chemopreventive approaches have been developed owing to the difficulties in target identification and validation [3]. Cell cycle arrest and apoptosis-inducing compounds have been shown to control cancer cell proliferation, which are the promising approaches to control tumorigenesis [4,5]. It has also been shown that reactive oxygen species (ROS) mediated DNA damage following treatment ß 2016 WILEY PERIODICALS, INC.

with chemotherapeutic agents including cisplatin, taxol, and radiation therapy, is an important trigger in the induction of apoptosis and cell death in a variety of cancer cells [6–10]. Hence, agents that are able to cause selective DNA damage and subsequent apoptosis in cancer cells will be promising therapeutic means to interfere with tumor growth and recurrence. Epidemiological studies suggest that dietary intake of vegetables and fruits may lower the risk of various types of malignancies including gastric cancer [11–15]. Naturally occurring flavonoids are a

Abbreviations: MRC I, mitochondrial respiratory complex I; MRC III, mitochondrial respiratory complex III; ROS, reactive oxygen species; FS, fisetin; NAC, N-acetyl cysteine; DCFDA, 20 ,70 dichlorofluorescein diacetate; PARP, poly(ADP-ribose) polymerase; CDKI, cyclin-dependent kinase inhibitor; R, rotenone; M, myxothiazol; DPI or D, diphenyleneiodonium chloride; BrdU, 5-bromo-20 -deoxyuridine; NP-40, nonyl phenoxypolyethoxylethanol-40; DHE, dihydroethidium. Grant sponsor: Central University of Gujarat; Grant sponsor: DSTPURSE; Grant sponsor: University Grant Commission *Correspondence to: 104 Cancer Biology Laboratory, School of Life Sciences, Jawaharlal Nehru University, New Delhi 110067, India. Received 17 January 2016; Revised 27 April 2016; Accepted 31 May 2016 DOI 10.1002/mc.22512 Published online in Wiley Online Library (wileyonlinelibrary.com).

2

SABARWAL ET AL.

group of low-molecular weight polyphenolic compounds synthesized by plants. Many of these compounds exhibit a wide range of biological activities, including antioxidant, anti-inflammatory, and antitumor effects [16–18]. Fisetin is a bioactive polyphenolic flavonoid, commonly found in many fruits and vegetables such as strawberries, apples, onions, and cucumbers [19]. Fisetin has attracted a great deal of research interest because of its remarkable anticancer effects against various cancer types [11–15]. Fisetin can suppress proliferation of cancer cells by causing cell cycle arrest and induction of apoptosis [20–23]. Elucidation of the mechanism of cellular responses to fisetin such as cell cycle arrest and apoptosis induction has been the topic of intense research in the past several years. For instance, fisetin caused decrease in the levels of cyclin D1, E, CDK 2, 4, and 6, and increase in cyclin dependent kinase inhibitors (CDKI) Cip1/ p21 and Kip1/p27 in prostate cancer cells [24]. Fisetininduced apoptosis in prostate cancer cells was correlated with the cleavage of caspases and simultaneous decrease in antiapoptotic protein Bcl-2 [24]. However, its anticancer activity and mechanisms against gastric cancer have not been studied yet. In the present study, we have investigated the effect of fisetin (Figure 1a) on cell viability, cell cycle, apoptosis, and associated molecular alterations in gastric cancer cell lines, AGS, and SNU-1. For the first time, we demonstrate that fisetin treatment inhibits complex I of the mitochondrial respiratory chain (MRC) in gastric cancer cells to trigger the generation of ROS, which causes DNA damage and apoptosis. These observations suggest that fisetin could be a potential cancer chemopreventive agent against gastric cancer. MATERIALS AND METHODS Chemicals and Reagents Fisetin (3,30 ,40 ,7-tetrahydroxyflavone), propidium iodide, saponin, bromophenol blue, ethylenediamine tetra acetic acid (EDTA), ethylene glycol tetra acetic acid (EGTA), and JC-1 dye, 20 ,70 -Dichlorofluorescin diacetate, rotenone, diphenyleneiodonium chloride (DPI), myxothiazol, N-Acetyl-L-cysteine (NAC), triton X-100, NP-40, aprotinin, and molecular biology grade DMSO were from Sigma–Aldrich (St. Louis, MO). MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide], disodium hydrogen phosphate (Na2HPO4), monopotassium phosphate (KH2PO4), were procured from Sigma–Aldrich. Ethanol was from Merck Biosciences (Darmstadt, Germany). RNase A was procured from Qiagen (Hilden, Germany). FITC Annexin-V apoptosis detection kit was purchased from BD Biosciences (San Jose, CA). MitoSox was purchased from molecular probes, Invitrogen (Carlsbad, CA). Complete cocktails of protease and phosphatase inhibitors were from Roche Molecular Biochemicals (Indianapolis, IN). PVDF membrane was from BD biosciences and luminata crescendo western HRP substrate (ECL) detection Molecular Carcinogenesis

system was from Merck Millipore (Billerica, MA). Cell proliferation ELISA BrdU kit was from Roche Diagnostics (Indianapolis, IN). DCFDA (20 ,70 -Dichlorofluorescin diacetate), and DHE (Dihydroethidium) were from Sigma (St. Louis, MO), MitoSOX was from Invitrogen, and Griess Reagent was from Thermo Fisher Scientific (Waltham, MA). All other chemicals and reagents were of highest purity grade. Cell Culture and Treatments The human AGS, SNU-1, and FHs74int cell lines were purchased from American Type Culture Collection (Manassas, VA). Human gastric epithelial carcinoma AGS cells (adherent) were maintained in Ham’s F-12 nutrient media. Human gastric epithelial carcinoma SNU-1 cells (non-adherent) were maintained in RPMI 1640 medium containing 2 mM L-glutamine, 10 mM HEPES, 1 mM sodium pyruvate, 4500 mg/L glucose, 1500 mg/L sodium bicarbonate, and supplemented with 10% fetal bovine serum (Gibco Life Technologies, Grand Island, NY, Invitrogen), 100 units/ml penicillin and streptomycin solution at 378C in humidified 5% CO2 incubator (Thermo Fisher Scientific). Non-neoplastic intestinal epithelial and adherent FHs74int cells were maintained in HybriCare medium (ATCC), containing 10 ng/ml EGF and 10% fetal bovine serum. The cells were grown as adherent monolayers or suspension cultures and fed every alternate day and passaged when cells reached 80% confluency [25]. We used 25–100 mM concentrations of fisetin to screen the effective doses of fisetin against gastric cancer cells, which we have used in our earlier studies. For cell proliferation studies, 100 mM was found to be cytotoxic to cancer as well as normal cells, and therefore omitted for further studies. And 50 mM concentration of fisetin was effective against both the cell lines, and thus used for mechanistic studies. Cultures were treated with fisetin (25, 50, 75, and 100 mM, final concentrations in medium) dissolved in dimethyl sulfoxide (DMSO) for different time points (24–72 h) for cell growth and 50 mM was used (0–48 h) for cell cycle analysis. In NAC-fisetin combination studies, cells were treated with NAC (10 mM) 5 h prior to fisetin treatment. In ROS inhibitor study, cells were treated with rotenone (0.5 mM), DPI (10 mM), and myxothiazol (0.6 mM) 2 h prior to fisetin treatment (50 mM) for indicated time point. The final concentration of DMSO was 0.1% v/v in each treatment. MTT Assay The assay detects the reduction of MTT [3-(4,5dimethylthiazolyl)-2,5-diphenyl-tetrazolium bromide] by mitochondrial dehydrogenases to blue formazan product, which reflects the normal function of mitochondria and hence the measurement of cytotoxicity or cell viability. Cells were seeded in 96-well plates at a density of 10 000 cells/well for 24 h and then treated with fisetin

FISETIN TARGETS MITOCHONDRIA FOR ROS-MEDIATED APOPTOSIS

3

Figure 1. Effect of fisetin on cell proliferation and death in human gastric cancer cells. AGS, SNU-1 (gastric cancer), and FHs74int (nonneoplastic) cells were treated with DMSO (control) or different concentrations of fisetin (25–100 mM FS) for 24, 48, and 72 h. At the end of each treatment time, adherent and non-adherent cells were collected and processed for determination of total cell number and dead cells as mentioned in Materials and Methods. (a) Chemical

structure of fisetin, (b) BrdU assay for AGS cells, (c) total cell number and dead cells of AGS cells, (d) total cell number and dead cells of SNU-1 cells, and (e) total cell number and dead cells of FHs74int cells. Columns, mean of three independent experiments; bars, s.e.m. Data points are the means s.e.m. of three experiments. P < 0.001 ( ), P < 0.01 ( ), P < 0.05 ( ). The P-value is determined by comparing each treatment with control group.

for various time points as indicated in the figure. Before the end of the treatments, 100 ml of 0.5 mg/ml MTT was added to each well and incubated at 378C for 4 h. After incubation the media were carefully

aspirated, and the formazan crystals were dissolved in 100 ml DMSO. Absorbance was measured at 550 nm by microplate reader (Synergy H1 Hybrid Reader, Biotek, Winooski, VT).

Molecular Carcinogenesis

4

SABARWAL ET AL.

Cell Growth and Death Assay Cells (1 105 per 60 mm culture plate) were seeded and next day treated with either DMSO or indicated concentrations of fisetin for 24, 48, and 72 h. For the data in Figure 8A, AGS cells were pre-treated with 0.5 mM ascorbic acid and/or 50 mM fisetin for 24 h. After each treatment time, cells were harvested by brief trypsinization and cell pellet was suspended in PBS and trypan blue dye was added (Sigma, Merck) and was counted by hemocytometer using a phase contrast microscope (Zeiss, Germany). Live cells were bright and unstained, whereas dead cells appeared blue. Total cells were obtained by adding both viable and non-viable cells. Percent dead cells were calculated by dividing dead cells by total number of cells multiplied by 100. Counting was done in duplicate for each triplicate sample and experiment was repeated at least three times. Cell Proliferation Assay The growth suppressive effect of fisetin on DNA synthesis which is correlated to cell proliferation was measured using a colorimetric bromodeoxyuridine (BrdU)-based Cell Proliferation ELISA kit (Roche Molecular Biochemicals, Mannheim, Germany) as per the manufacturer’s instructions. Briefly, 5000 cells/well in 96-well plate were plated and next day treated with different concentrations of fisetin (0–100 mM) and then incubated with BrdU solution and then with peroxidase-conjugated anti-BrdU antibody followed by substrate tetramethylbenzidine for 30 min at room temperature and fluorescence was detected at 370 (reference wavelength 492 nm) using microplate reader (Synergy H1 Hybrid Reader, BioTek).

1 ml of fresh cell culture medium, 2.5 ml JC-1 dye (stock 5 mg/ml) was added and vortexed for 10–15 s and incubated for 10 min in dark at 378C. Cells were washed twice with PBS and subsequently analyzed with flow cytometer [19]. Data was analyzed as JC-1 monomer/dimer ratio. Apoptosis Assay To quantify apoptosis, AGS and SNU-1 cells were seeded and treated similarly as for the cell growth assay. Cells were harvested and processed for apoptosis assay using FITC Annexin V Apoptosis detection kit 1 from BD Biosciences, following step-by-step protocol prescribed by the manufacturer and analyzed by flow cytometry. Briefly, at the end of treatment, both floating and attached cells were collected, and subjected to annexin V and PI staining. Flow cytometry was performed within 30 min using flow cytometer to score annexin V and/or PI positive cells. Lysate Preparation and Immunoblot Analysis

AGS and SNU-1 cells were seeded as for cell growth assay in complete medium. After 24 h, complete medium was replaced with fresh serum-free medium and cells were starved for 24 h. These synchronized cells were released using media containing serum with DMSO (control) or 50 mM fisetin and cell cycle analysis was done as a function of time up to 48 h. At the end of each treatment, total cells were collected and processed for cell cycle analysis as reported recently [26]. Briefly, cells were suspended in 0.5 ml of saponin/PI solution (0.3% saponin [w/v], 25 mg/ml PI [w/v], 0.1 mM EDTA, and 10 mg/ml RNase A [w/v] in PBS) and incubated for 4 h at 48C in dark. Cell cycle distribution was then analyzed by flow cytometry using FACS Aria III flow cytometer (BD Biosciences) [27].

Cells were treated at 70% confluency with fisetin at 25, 50, and 75 mM concentrations for desired time points. Cell lysates were prepared in non-denaturing lysis buffer (10 mM Tris–HCl, pH 7.4, 150 mM NaCl, 1% TritonX-100, 1 mM EDTA, 1 mM EGTA, 0.3 mM phenyl methyl sulfonyl fluoride, 0.2 mM sodium orthovanadate, 0.5% NP-40, 5 U/ml aprotinin) as reported earlier [4]. Protein concentrations in lysates were determined using Bradford reagent from Bio-Rad laboratories (Hercules, CA). The primary antibody for Actin, Bax, Bcl-2, cyclin E, and cleaved PARP were from BD Biosciences (San Diego, CA). Primary antibodies to Cip1/p21, Kip1/p27, survivin, cleaved caspase 3, phospho-H2A.X (S139), and secondary anti-rabbit and anti-mouse antibodies were from Cell Signaling (Beverly, MA). Primary antibodies for CDK2, CDK4, CDK6, Cyclin D1, total p53, and phospho-p53 (S15) were from Santa Cruz Biotechnology (Santa Cruz, CA) and Mdm2 was from Abcam (Cambridge, MA). Cell lysates (30–50 mg/sample) were denatured in SDS–PAGE sample buffer on boiling water bath for 5 min. Samples were loaded onto 8–10% denaturing SDS–PAGE gels and proteins resolved at constant voltage. Proteins from the gel were transferred onto PVDF membrane using transfer assembly (Bio-Rad). Membranes were blocked in blocking buffer for 1 h and incubated with specific primary antibody followed by appropriate HRP-linked secondary antibody and processed for ECL detection. Bands on x-ray films were scanned using high-resolution scanner [28].

JC-1 Staining for Mitochondrial Membrane Potential

Comet Assay

To determine the mitochondrial membrane potential, AGS and SNU-1 cells were seeded and treated for 24 and 48 h similarly as done for cell growth assay. Cells were harvested and 2 105 cells were taken in

The Comet assay was carried out as described [29]. Briefly, the cells were suspended in 1% low melting point agarose at a concentration of 1 104 cells/ml and applied to the surface of a microscope slide. The

Flow Cytometric Analysis for Cell Cycle Phase Distribution



slides were immersed in lysis buffer (2.5 M NaCl, 10 mM Tris base, pH 10.5–11, 100 mM EDTA, and 1% triton X 100 and 1% DMSO added freshly before use) for 1 h, and then slides were immersed with unwinding buffer (0.3 M NaOH, 0.5 M EDTA) for 20 min, both at 48C. Slides were then soaked in neutralization buffer (500 mM Tris–HCl, pH 8.0) for 20 min. After slides were put to horizontal gel electrophoresis, they were washed and stained with 1 mg/ml ethidium bromide. The cells were analyzed with fluorescence microscope and the proportion of the cells containing damaged DNA/strand breaks/appearance of tails as the comet cells was scored. DNA-damage-induced comet cells were also analyzed with image J software. Each cell was assigned a value of 0 (without tail) to maximum (almost all DNA as tail), and the total score of 100 cells were represented the DNA damage level, which were in the range between 0 and 400 “arbitrary units.” DCF-DA Staining Cells were cultured at 12 103 cells/well in 96-well plate. After 24 h, spent media was aspirated and wells were washed with Krebs-ringer bicarbonate solution and incubated with 10 mM freshly prepared DCF-DA solution in dark. After 30 min, DCF-DA was removed and cells were washed twice with Krebs-ringer bicarbonate solution. Five hours pre-treatment of NAC was done to block ROS for wells which were later treated with H2O2 (1 mM) and fisetin (25–50 mM). Finally, the fluorescence was measured at different time-points (0, 1, 3, 6 h) using microplate reader with excitation filter set at 485 nm and emission filter set at 530 nm. MitoSox Staining Cells (1 104 cells/well) in six-well plate were seeded and after 24 h spent media was removed and cells were treated with either DMSO (control) or 50 mM fisetin in fresh media. After indicated time points (3–12 h), cells were incubated with MitoSox (5 mM, Molecular Probes) for 1 h for the detection of mitochondrial superoxide generation at 378C and analyzed for fluorescence intensity by flow cytometry (BD Biosciences, San Jose, CA). DHE Staining Cells were seeded in 6-well (1 104 cells/well) plate for microscopy and 96-well (12 103) plate for fluorometry. Five hours pre-treatment of NAC was done to block ROS for wells which were later treated with fisetin (50 mM) for 6 h. Then cells were incubated with DHE (2.5 mM) for 20 min followed by fluorometry detection at excitation and emission at 535 and 635 nm, respectively using microplate reader. Six-well plate was treated in similar manner and after incubation cells were detached with brief trypsinization and transferred to glass slides for microscopy using Carl Zeiss fluorescent microscope. Molecular Carcinogenesis

5

Nitric Oxide Measurement Nitric oxide produced by AGS cells was assayed by the measurement of nitrite, a stable NO oxidation product. Cells (5 105 cells/well) were plated in sixwell tissue culture plates and incubated at 378C for 24 h in a 5% CO2 humidified incubator. Five-hour pretreatment of NAC was done to block RNS for wells which were treated with fisetin (50 mM). After 24 h of incubation with fisetin, spent culture medium aspirated from wells of culture plate (100 ml) and was mixed with 100 ml of Griess reagent, and the absorbance of the mixture was measured in a microplate reader at 540 nm. Concentration of nitrite in mM was calculated using standard curve developed with known concentration of sodium nitrite. Statistical Analysis Statistical significance of differences between control and treated samples were calculated using Student’s t-test (Graphpad Prism version 5). P-value of less than 0.05 were considered significant. ANOVA was followed by Bonferroni’s Test for Multiple Comparisons, for Figure 7b. RESULTS Fisetin Inhibits Cell Proliferation of Human Gastric Cancer AGS Cells To explore whether fisetin treatment has inhibitory effect on DNA synthesis in gastric cancer cells AGC cells were used for a colorimetric BrdU proliferation assay. The principle of this assay involves the incorporation of pyridine analogue BrdU (instead of thymidine) into the newly synthesized DNA of proliferative cell. After its incorporation, BrdU is detected by immunoassay. Cells were treated with different concentrations of fisetin (25–100 mM) for 24 and 48 h, and a time- and dose-dependent decrease was observed in cell proliferation of AGS cells (Figure 1b). Fisetin Inhibits Growth and Viability of Human Gastric Cancer AGS and SNU-1 Cells Next, we examined the effect of fisetin on the growth and viability of human gastric cancer AGS and SNU-1 cells. Fisetin (25–100 mM) treatment for 24 h reduced total cell number by 4–64% (P < 0.05–0.001) of AGS cells (Figure 1c left). A further decrease in cell growth of 32–79% and 24–85% at similar fisetin concentrations were also observed following prolonged treatment durations of 48 and 72 h, respectively (P < 0.001) (Figure 1c left). Under similar treatment conditions, fisetin increased the dead cell population by 1.5–3.0-fold after 24–72 h of treatment, respectively (Figure 1c right). The observed lesser number of dead cells at higher concentrations of fisetin and at later time points were probably due to the conversion of dead cells into debris, which were observed in these treatments. These observations

6

SABARWAL ET AL.

suggested that 50–100 mM fisetin was most effective in inducing cell death in these cancer cells. Similarly in SNU-1 cells, fisetin reduced total cell number by 23–45%, 65–79%, and 83–93% after 24–72 h of treatment (Figure 1d left) (P < 0.01–0.001) and increased the dead cell population by 1.5–9.0-fold after 24–72 h of treatment, respectively (Figure 1d right). MTT assay of AGS cells further confirmed the decrease in cell viability in time- and dose-dependent fashion (Supplementary Figure S1). Next, we studied the effect of fisetin on growth and viability of non-neoplastic FHs74int, intestinal cells. Fisetin was less effective in decreasing the cell viability and proliferation of these cells, as compared to cancer cells. In FHs74int cells, 48 h of fisetin treatment did not decreased cell number (Figure 1e left) compared to 79% (P < 0.001) in gastric cancer cell lines. Fisetin treatment increased the dead cell population by 1.3-fold (P < 0.05) after 24 h, at highest concentration of 100 mM. After 48 h of treatment only higher doses (75 and 100 mM) caused an increase in dead cells which was not statistically significant (Figure 1e right). Overall, a relatively selective effect of fisetin was observed against gastric cancer cells. Fisetin-Induced Strong G1 Phase Cell Cycle Arrest in Gastric Cancer Cells To investigate the growth inhibitory mechanisms of fisetin on gastric cancer cells, cell cycle distribution analysis of AGS and SNU-1 cells were performed. Cells were synchronised by serum-starvation for 24 h which induced G0/G1 phase arrest. After synchronization, cells were released using medium containing serum without or with fisetin (50 mM), and cell cycle analysis was done as a function of time from 0 to 48 h. As shown in Figure 2, untreated AGS and SNU-1 cells started entering S phase after 12 and 9 h of the release, respectively, and passed through subsequent phases of the cell cycle. However, fisetin-treated AGS cells remained arrested in G1 phase even after 48 h of the treatment (Figure 2). These results demonstrated that fisetin induces a G1 phase arrest in gastric cancer cells. Relatively, AGS cells showed more prominent arrest whereas SNU-1 cells started relieving from the arrest after 36 h of fisetin treatment. Next, we examined the molecular alterations associated with G1 phase arrest in gastric cancer cells. Effect of Fisetin on the Protein Levels of CDK Inhibitors, Cyclins, and CDKs We analyzed G1 phase cell cycle regulatory molecules which could be altered by the fisetin treatment (Figure 3). Fisetin increased the level of cyclindependent kinase inhibitor (CDKI) Cip1/p21 in AGS cells at 25–50 mM, whereas decreased its level at 75 mM concentration (Figure 3a left). However, a dose- and time-dependent increase in the level of Cip1/p21 was observed in SNU-1 cells (Figure 3a right). In case of Kip1/p27, there was an initial increase at 25 mM fisetin Molecular Carcinogenesis

followed by its reduction at higher concentrations (Figure 3a). The Cip/Kip proteins are upregulated by several pathways, both at the transcriptional and post-transcriptional levels and are known to be extremely short-lived. Their metabolic stability and degradation through proteolytic pathways are determining factors that decide the fate of the cells through the cell cycle [30]. We also found an up-regulation of total p53 level in time- and dose-dependent manner in both the cell lines; however, at highest concentration (75 mM) for 48 h treatment, p53 was found to be down-regulated in SNU-1 cells. Interestingly, we also found activation of p53 by its phosphorylation at S15 position in SNU-1 cells, indicating the likely involvement of DNA damage in these cells (Figure 3b). Simultaneously, a decrease in the expression of Mdm2, a negative regulator of p53 was also observed which further strengthened our results. Fisetin decreased the expression of protein levels of G1 regulatory cyclins and CDKs including CDK2, CDK4, cyclin D1, and cyclin E in AGS cells, and CDK2, CDK4, and cyclin D1 in SNU-1 cells (Figure 3c). These results suggested that fisetin induces G1 phase arrest in gastric cancer cells by modulating the expression levels of CDK-Cyclin-CDKI as well as p53. Fisetin Induces Apoptosis and Mitochondrial Membrane Depolarization in Human Gastric Cancer Cells An increase in dead cells by fisetin was observed, hence, we examined whether fisetin-induced death was via induction of apoptosis. Consistent with the cell death effect, 25–75 mM fisetin treatment for 24 h showed a dose-dependent increase in the apoptotic cells (Figure 4a). Fisetin increased apoptotic cells by 11–23% (P < 0.001) at 24 h and 9–25% (P < 0.05–0.001) at 48 h of treatment in AGS cells (Figure 4b left). In SNU-1 cells, fisetin (25–50 mM) caused upto four-fold increase in apoptotic cells (P < 0.001) at 24 h of treatment (Figure 4b right). However, 75 mM fisetin increased apoptotic cells by 3.5% (P < 0.001) and also showed necrotic cells and debris (data not shown). The collapse of mitochondrial membrane potential barrier is considered as one of the earliest signs for the induction of apoptosis [31]. We next evaluated the effect of fisetin on mitochondrial membrane depolarization using JC-1 cationic dye. In healthy or live cells within intact mitochondria, JC-1 forms dimer, whereas in apoptotic cells due to loss of mitochondrial potential, it comes out of mitochondria into the cytoplasm and dissociates into monomers [32]. After treatment with 25–75 mM fisetin for 24 and 48 h of AGS cells showed (Figure 4c left) dissociation of JC-1 from dimer to monomer form. The monomer/dimer ratio increased by up to four-fold (P < 0.05–0.001) in presence of fisetin at 25–75 mM after 48 h of treatment (Figure 4c left). In SNU-1 cells, 25–75 mM fisetin for 24–48 h increased monomer/dimer ratio up to 4–20-fold (P < 0.001, Figure 4c right). These findings suggested the role of


7

Figure 2. Effect of fisetin on cell cycle distribution in human gastric cancer cells. (a) AGS cells in G1, S, and G2/M phases, (b) SNU-1 cells in G1, S, and G2/M phases. Cells were serum-starved to synchronize them in G1 phase and then released with addition of serum and treated with either DMSO control for 24 h or 50 mM fisetin (FS) for indicated time points. At the end of these treatments, adherent and

non-adherent cells were collected and incubated overnight with saponin/PI solution at 48C as detailed in Materials and Methods. The percentage of cells in the different phases of the cell cycle was determined by flow cytometry and data points are presented as mean of triplicates which was reproducible in additional two experiments.

mitochondria-mediated apoptosis as one of the mechanisms of cell death induced by fisetin.

level was strongly decreased by fisetin treatment in a dose-dependent fashion in SNU-1 cells (Figure 5b). We could not detect Bcl-2 in AGS cells. However, a significant increase in the ratio of Bax to Bcl-2 was observed after treatment with fisetin in SNU-1 cells (Figure 5c). Additionally, as fisetin has caused upregulation of total p53 and its activation by phosphorylation at S15 position, indicating the likely involvement of DNA damage. In case of DNA strand breaks, usually the phosphorylation of the histone H2AX takes place. It is phosphorylated by kinases such as ataxia telangiectasia mutated (ATM) and ATM-Rad3-related (ATR) in the PI3K pathway [33]. This newly phosphorylated protein, gamma-H2A.X, is the first step in recruiting and localizing DNA repair

Effect of Fisetin on Apoptosis Regulating Proteins Furthermore, we investigated the expression levels of proteins involved in apoptosis and found the increase in cleavages of caspase 3 and PARP in both the gastric cancer cell lines at higher doses and in AGS cells at later time point. However, the increase in expression level was more prominent in SNU-1 cells (Figure 5a and b). Bax and Bcl-2 play significant role in apoptosis induction, the immunoblot analysis showed an increase in the level of Bax in timeand dose-dependent fashion in AGS cells, with no considerable change in SNU-1 cells, whereas Bcl-2 Molecular Carcinogenesis

8

SABARWAL ET AL.

Figure 3. Effect of fisetin on cell cycle regulatory proteins in gastric cancer cells. AGS and SNU-1 cells were treated with the indicated doses of fisetin (FS) for 24 and 48 h. At the end of the treatments, lysates were analyzed by immunoblotting using specific primary antibodies for (a) p21 and p27, (b) p-P53(S15), P53 total, Mdm2 (c) CDK2, 4 and 6, cyclin D1 and E, followed by detection with HRP-labeled appropriate secondary antibodies as mentioned in Materials and Methods. b-Actin was probed after stripping the membrane as protein loading control.

proteins. As anticipated, fisetin increased the phosphorylation of gamma-H2A.X S139 in both the cell lines (Figure 5a and c). Next, we assessed the effect of fisetin on the levels of survivin. Overexpression of survivin neutralizes apoptosis of B lymphocytes and proved its role in apoptosis inhibition which is a general feature of neoplasia and thus survivin is Molecular Carcinogenesis

regarded as a potential new target for apoptosis-based therapy in cancer [34,35]. We observed a decrease in the levels of survivin by fisetin in gastric cancer cells which further strengthens our results that fisetin decreases antiapoptotic proteins to promote apoptosis. Moreover, high level of phosphorylation of gamma-H2A.X S139 was found at 24–48 h, but when


9

Figure 4. Effect of fisetin on apoptotic death in human gastric carcinoma cells. Cells were treated with either DMSO (Control) or 25–75 mM fisetin (FS) for 24 and 48 h. At the end of treatments, cells were harvested and stained with annexin V and PI as detailed in Materials and Methods. (a and b) Apoptotic cells were analyzed by flow cytometry as detailed in Materials and Methods. (c) Effect of fisetin on mitochondrial membrane potential in human gastric carcinoma AGS

and SNU-1 cells. Live cells with intact mitochondrial membrane potential and dead cells with depolarized mitochondrial membrane potential was measured by JC-1 staining and analyzed by flow cytometry as described in Materials and Methods. Columns, mean of three independent treatments; bars, s.e.m. Data points are the means s.e.m. of three experiments. P < 0.001 ( ), P < 0.01 ( ), P < 0.05 ( ). The P-value is determined by comparing each treatment with control group.

we performed time kinetics for gamma-H2A.X S139 phosphorylation and cleaved PARP, we observed that the initiation of DNA damage might have occurred as early as 3 h and cells started entering into apoptosis after 12 h of the fisetin treatment (Figure 5d).

proapoptotic or cytoprotective effects depending on the nature and extent of the damage [36]. Since, fisetin increased p53 levels in both the cell lines indicating its likely involvement in apoptosis. Pifithrin a, an inhibitor of p53 was employed to further investigate the relevance of p53 function in AGS cells in response to fisetin treatment. Under p53 inhibitory condition, we did cell viability and Annexin V–FITC assay. Under p53 inhibitory condition total number of cells were further decreased and dead cells were increased by fisetin

Role of p53 in Fisetin-Mediated Apoptosis of Gastric Cancer Cells As a consequence of multiple functions of p53, its activation in response to cytotoxic stress may have Molecular Carcinogenesis

10

SABARWAL ET AL.

Figure 5. Effect of fisetin on apoptotic regulatory proteins in gastric cancer cells. AGS and SNU-1 cells were treated with different doses of fisetin (FS) for indicated time points. (a and b) At the end of the treatments, lysates were analyzed by immunoblotting using specific primary antibodies for cleaved caspase-3, cleaved PARP, p-H2A. X(S139), survivin, Bax, and Bcl-2 followed by detection with

HRP-labeled appropriate secondary antibodies as mentioned in Materials and Methods. b-Actin was probed after stripping the membrane as protein loading control. (c) Bax/Bcl2 ratio at indicated time points and doses of fisetin treatment. (d) Time kinetics western blotting for p-H2A.X(S139) and cleaved-PARP at indicated concentration and time point.

treatment (decreased cell number from 48% at 50 mM fisetin to 71% [P < 0.001] at 75 mM pifithrin a þ 50 mM fisetin and increased cell death from 9% to 13% in same treatments, Figure 6a). Under similar treatment conditions number of apoptotic cells also increased from 15% to 20% (Figure 6b). Pre-treatment of pifithrin a (75 mM) for 2 h before fisetin (50 mM) treatment was found to be appropriate to inhibit p53 expression which was upregulated by fisetin in AGS cells at 24 h (Figure 6c). Inhibitor remained in the media till the end of experiment. These results suggested that fisetininduced p53 levels has no role in cell death rather it is independent of p53 induction. We further explored the mechanism of cell death caused by fisetin.

to fisetin and the changes in DCF fluorescence was measured. Fisetin increased ROS generation in a dosedependent manner (Supplementary Figure S2) and it was blocked by an antioxidant, NAC (Figure 7a). To determine the source of ROS we further analyzed mitochondrial superoxide generation using MitoSox staining and flow cytometry analysis. We observed a significant increase in mitochondrial superoxide generation at 50 mM fisetin at 3 and 6 h; however, at 12 h ROS generation from mitochondria was decreased. The decrease in mitochondrial ROS might be due to the decrease in total healthy mitochondria in treated samples (Figure 7b). To find out the role of ROS in DNA damage, comet assay was performed in presence of fisetin alone and in combination with NAC (Figure 7c). Interestingly, NAC pre-treated cells did not show any comet tails (Figure 7c left) whereas fisetin alone showed comet tails of high intensity at 6 h of treatment which could happen only due to severe DNA damage. Next, we determined whether

Fisetin Induces ROS Formation and Caused DNA Damage in Gastric Cancer Cells Many studies have shown that oxidative stress has antitumor effects in cancer cells [37,38]. To study the effect of fisetin on ROS generation, cells were exposed Molecular Carcinogenesis


11

Figure 6. Role of p53 in fisetin-mediated death of gastric cancer cells. AGS cells were pre-treated with treated with pifithrin-a (Pif) and/or the indicated concentrations of fisetin (FS) for 24 h. (a) Cells were collected and processed for determination of total cell number and percent dead cells of AGS cells, (b) induction of apoptotic cell death by fisetin and/or pifithrin was assessed by Anexin V-FITC assay in AGS cells. (c) For the effective concentration of pifithrin, the lysates were analyzed by immunoblotting using specific primary

antibodies for p53 followed by detection with HRP-labeled appropriate secondary antibodies as mentioned in Materials and Methods. b-Actin was probed after stripping the membrane as protein loading control. Columns, mean of triplicates which was reproduced in additional two-three independent experiments; bars, s.e.m. Data points are the means s.e.m. of triplicates for each treatment. P < 0.001 ( ), P < 0.05 ( ). The P-value was determined by comparing each treatment with control group.

ROS generation is involved in fisetin-induced cell death. Fisetin-induced cell death was strongly prevented by NAC pre-treatment in both AGS (Figure 7d right) and SNU-1 cells (Supplementary Figure S2). These results suggested that fisetininduced cell death is caused by ROS generation. We further analyzed the levels of cellular superoxides by DHE staining and nitrite which were strongly increased by fisetin treatment and these effects were lost when the cells were pre-treated with NAC (Figure 7e and f). We also analyzed the effects on apoptosis and DNA damage marker, cleaved-PARP, and gamma-H2A.X S139 phosphorylation, respectively in similar treatments. The levels of both were increased by fisetin treatment; however, pretreatment with NAC for 5 h inhibited the increase in the levels of both the proteins (Figure 7g), and thus suggesting that fisetin generated ROS caused DNA damage and apoptosis in gastric cancer cells. The fisetin-caused generation of ROS was further proved

by using another known antioxidant, ascorbic acid, the pre-treatment of which reversed the growth inhibitory and cell death inducing effects of fisetin (Figure 8a and b).


Fisetin Generates ROS Through MRC I Subcellular organelles, including mitochondria are major source of cellular ROS, which are generated due to incomplete reduction of oxygen during normal oxidative phosphorylation [39–41]. To test whether fisetin-induced ROS generation was mitochondrialderived, we investigated the effect of mitochondrial respiratory complex (MRC) inhibitors namely, rotenone and myxothiazol on fisetin-mediated ROS generation. Rotenone is an inhibitor of MRC I, and myxothiazol is an inhibitor of MRC III. Our results suggest that (Figure 8c) (Supplementary Figure S4) fisetin-mediated increase in DCF-DA fluorescence was significantly inhibited in the presence of rotenone. However, pre-treatment with myxothiazol conferred

12

SABARWAL ET AL.

Figure 7. Detection of ROS using carboxy-H2DCFDA dye and MitoSox. (a) AGS cells were treated as indicated in the figure and DCF fluorescence were measured. (b) Levels of mitochondrial superoxide was quantified using MitoSox by flow cytometry and analyzed by ANOVA followed by Bonferroni's Multiple Comparison Test. (c) The formation of comets (fragmented DNA) are shown at 400 magnifications and in a magnified view with quantittative data. (d) AGS cells were collected and processed for determination of total cell number and dead cells employing trypan blue assay. (e) DHE and (f) nitrite assays were done in AGS cells as detailed in


Materials and Methods. (g) Fisetin-mediated ROS generation and DNA damage was inhibited in the presence NAC. AGS cells were treated with Fisetin (50 mM) after 6 h lysates were collected and status of H2A.X(S139) and cleaved PARP were determined by western blot analysis. Columns, mean of three independent treatments; bars, s.e.m.; the results were reproducible in two additional experiments. Data points are the means s.e.m. of triplcates. P < 0.001 ( ), P < 0.01 ( ), P < 0.05 ( ). The P-value is determined by comparing each treatment with control group or indicated groups.


Figure 8. Proposed mechanism of fisetin-induced apoptosis in human gastric cancer cells. (a and b) Effect of pre-treatment of ascorbic acid on cell proliferation and death of AGS cells. Cells were pre-treated with 0.5 mM ascorbic acid and/or 50 mM fisetin for 24 h, and number of total and dead cells were counted using trypan blue with hemocytometer. (c) Effect of pre-treatment with rotenone and myxothiazol on fisetin-mediated ROS generation in AGS cells after 6 h. (d) Effect of pre-treatment with DPI and myxothiazol on fisetinmediated ROS generation in AGS cells after 6 h. The cells were pretreated for 2 h with either DMSO (control), 0.5 mM rotenone (R), 10 mM DPI (D), or 0.6 mM myxothiazol (M). The cells were then either left


13

untreated (DMSO, rotenone, DPI, and myxothiazol alone control groups) or exposed to 50 mM fisetin for 6 h for detection of ROS generation. (e) Proposed mechanism of fisetin-mediated ROS generation from mitochondrial respiratory chain complex I that leads to DNA damage. In response to DNA damage, cells get arrested in G1 phase and apoptosis induction occurs in gastric cancer cells. Results are mean s.e.m. (n ¼ 3); significantly different (P < 0.05) compared with DMSO-treated control. Columns, mean of three independent wells; bars, s.e.m.; the results were reproducible in two additional experiments. The P-value was determined by comparing each treatment with control or indicated groups.

14

SABARWAL ET AL.

no protection against fisetin-mediated ROS generation. To further confirm the involvement of MRC complex I, we used another inhibitor of complex I, DPI, to inhibit the generation of ROS, and observed that fisetin-caused ROS generation was attenuated in the presence of DPI (Figure 8d) (Supplementary Figure S5). Altogether, our data suggested that ROS upon treatment of human gastric cancer cells with fisetin is generated from mitochondria and mostly through MRC complex I. However, further studies are needed to confirm the role and mechanisms of MRC 1 in generation of ROS by fisetin. DISCUSSION The central findings of the present study are that fisetin (a) inhibits growth and survival of human gastric cancer cells; (b) shows selectivity for these effects as normal cells are very less affected; (c) induces G1 phase arrest by modulation of CDK-cyclin-CDKI levels; (d) dissipates mitochondrial membrane potential and up-regulates proapoptotic molecules and down-regulates anti-apoptotic molecules for apoptosis induction; (e) induces p53 level; however, it has not role in cell death; and (f) causes ROS generation likely through MRC I and leads to DNA damage and apoptosis in gastric cancer cells (Figure 8e). Since, fisetin is a dietary phytochemical and relatively safer, it has strong potential for cancer chemopreventive activity against gastric cancer through dietary intervention. Fisetin is shown to inhibit proliferation of many types of cancer cells [11–15]. For the first time we observed that it inhibited the proliferation of human gastric cancer cells. This effect was rather selective as compared to normal intestinal cells. The growth inhibition was accompanied by the changes in cell cycle regulatory molecules that led to G1 phase arrest in cancer cells. The more prominent changes observed were the increased levels of cip1/p21 and p53 and decrease in CDK2, CDK4, and cyclin D1 protein levels. CDK, the catalytic subunit and cyclin, the regulatory subunit, are required to form a complex of an active kinase to transit the cell through the cell cycle check-points. Even this active kinase complex can be inhibited by the physical interaction of CDKI such as cip1/p21 [42]. Such molecular alterations are very well known to induce G1 phase arrest [43]. Our results suggest that fisetin did not let the cells come out from the G1 phase to enter in the S phase for a longer time and thus cells remain arrested in G1 phase. This effect was likely mediated via the downregulation of G1 phase CDK and cyclins, and partly due to increase in CDKIs. However, at higher concentration of fisetin, mostly we observed a decrease in the level of CDKI which could be associated with the cell death or apoptosis. As sudden release of cells with damaged DNA from the G1/S phase could drive the cells for cell death or apoptosis. Molecular Carcinogenesis

Fisetin inhibited the survival of gastric cancer cells through inducing apoptosis by several fold. During mechanistic investigation, we observed that fisetininduced mitochondrial membrane depolarization and increased the levels of cleaved caspase-3 and PARP. These changes are characteristics of the role of mitochondrial apoptosis [44]. An increase in Bax and a decrease in survivin level in AGS cells, and a decrease in Bcl-2 and survivin levels in SNU-1, showing an overall increase in Bax/Bcl-2 ratio explained the apoptotic effect of fisetin in gastric cancer cells. It also indicated that different molecules could be targeted by fisetin to achieve the apoptosis in different gastric cancer cells. Surprisingly, we observed that fisetin-induced p53 level and its phosphorylation with decreased level of mdm-2 did not contribute to apoptosis. The presence of pifithrin-alpha, an inhibitor of p53, could not inhibit the fisetin-induced growth inhibition and apoptosis in gastric cancer cells. Further our study demonstrated that ROS acts as key signaling intermediates in fisetin-induced apoptosis in human gastric cancer cells. The ROS generation and cell death resulting from fisetin treatment were inhibited by pre-treatment of NAC suggesting that ROS assisted to initiate the process of cell death. ROS generation by fisetin has been documented previously in human U266 myeloma cell line [45]. Fisetininduced apoptosis was related to AMPK activation and ROS production in U266 cells; however, the source of ROS generation was not known. Fisetin caused an increase in levels of cleaved-PARP and a decrease in phosphorylation of mTOR [45]. Conversely, fisetin has also shown anti-oxidative properties in other cancer models, for instance in case of hepatocellular carcinoma (HCC) in rats, enhanced level of ROS was associated with aflatoxin-B1-induced hepatocellular carcinoma progression. A significant rise in ROS level was observed in the liver from the HCC rats as compared to the control. However, after fisetin treatment, the enhanced ROS level in HCC liver was seen to be decreased significantly to regain its normal value; fisetin also caused increase in total glutathione levels [46]. Hydrogen peroxide (H2O2) can induce damage to the cell by generating reactive oxygen species (ROS), resulting in DNA damage and cell death. Fisetin reduced the level of superoxide anion, hydroxyl radical in cell free system, and intracellular ROS generated by H2O2, additionally fisetin also protected against H2O2-induced lipid peroxidation [47]. Our study showed that fisetininduced cell death is partly reversed by ascorbic acid which is a potent antioxidant, and thus further indicating the involvement of ROS in fisetin-induced apoptosis. Although a possible contribution of ROS in apoptotic response to fisetin has been suggested [47], the mechanism of ROS generation was not clear. The present study provides convincing experimental evidence to indicate that ROS generation by fisetin in


human gastric cancer cells is mostly mitochondrial derived. Cellular superoxides and nitrites were also produced. The mitochondria-derived ROS by fisetin is supported by our findings including the fisetinmediated ROS generation is significantly attenuated in the presence of MRC complex I inhibitors, rotenone, and DPI but not in the presence of complex III inhibitor, myxothiazol (Figure 8a). The ROS scavenger, NAC, significantly blocked cell death caused by fisetin in both gastric cancer cell lines, AGS and SNU-1 cells (Figure 7d, Supplementary Figure S3). Further, NAC pre-treatment also blocked the activation of PARP and gamma-H2A.X S139 phosphorylation which suggested that fisetin-mediated DNA damage and subsequent apoptosis was likely triggered by ROS. Although further studies are needed to determine the precise mechanism of fisetin-mediated generation of ROS through complex I. Overall, for the first time our data suggested that fisetin caused ROS generation through mitochondrial complex I that leads to DNA damage in human gastric cancer cells (Figure 8c). In response to DNA damage, cells get arrested in G1 phase and subsequently mitochondrial apoptosis is triggered in these cells. Surprisingly, fisetin-induced p53 did not had any role in fisetin-caused apoptotic death of gastric cancer cells. ACKNOWLEDGMENTS The work was supported by the fund from the Central University of Gujarat and DST-PURSE, India, and AS is supported by a fellowship from University Grant Commission, New Delhi, India. REFERENCES 1. Ferlay J, Soerjomataram I, Dikshit R, et al. Cancer incidence and mortality worldwide: Sources, methods and major patterns in GLOBOCAN 2012. Int J Cancer 2015;136:E359–E386. 2. Bertuccio P, Chatenoud L, Levi F, et al. Recent patterns in gastric cancer: A global overview. Int J Cancer 2009;125: 666–673. 3. Delaunoit T. Latest developments and emerging treatment options in the management of stomach cancer. Cancer Manag Res 2011;3:257–266. 4. Tyagi A, Agarwal C, Harrison G, Glode LM, Agarwal R. Silibinin causes cell cycle arrest and apoptosis in human bladder transitional cell carcinoma cells by regulating CDKICDK-cyclin cascade, and caspase 3 and PARP cleavages. Carcinogenesis 2004;25:1711–1720. 5. Singh RP, Dhanalakshmi S, Agarwal C, Agarwal R. Silibinin strongly inhibits growth and survival of human endothelial cells via cell cycle arrest and downregulation of survivin, Akt and NF-kappaB: Implications for angioprevention and antiangiogenic therapy. Oncogene 2005;24:1188–1202. 6. Xiao D, Powolny AA, Singh SV. Benzyl isothiocyanate targets mitochondrial respiratory chain to trigger reactive oxygen species-dependent apoptosis in human breast cancer cells. J Biol Chem 2008;283:30151–30163. 7. Cao J, Jia L, Zhou HM, Liu Y, Zhong LF. Mitochondrial and nuclear DNA damage induced by curcumin in human hepatoma G2 cells. Toxicol Sci 2006;91:476–483. 8. Sakao K, Singh SV. D,L-sulforaphane-induced apoptosis in human breast cancer cells is regulated by the adapter protein p66Shc. J Cell Biochem 2012;113:599–610.


15

9. Alexandre J, Batteux F, Nicco C, et al. Accumulation of hydrogen peroxide is an early and crucial step for paclitaxelinduced cancer cell death both in vitro and in vivo. Int J Cancer 2006;119:41–48. 10. Bickers DR, Athar M. Oxidative stress in the pathogenesis of skin disease. J Invest Dermatol 2006;126:2565–2575. 11. Tsugane S, Sasazuki S. Diet and the risk of gastric cancer: Review of epidemiological evidence. Gastric Cancer 2007;10: 75–83. 12. Cohen JH, Kristal AR, Stanford JL. Fruit and vegetable intakes and prostate cancer risk. J Natl Cancer Inst 2000;92:61–68. 13. Li Y, Fang H, Xu W. Recent advance in the research of flavonoids as anticancer agents. Mini Rev Med Chem 2007;7:663–678. 14. Kimira M, Arai Y, Shimoi K, Watanabe S. Japanese intake of flavonoids and isoflavonoids from foods. J Epidemiol 1998;8: 168–175. 15. Woo HD, Lee J, Choi IJ, et al. Dietary flavonoids and gastric cancer risk in a Korean population. Nutrients 2014;6: 4961–4973. 16. Middleton E, Jr., Kandaswami C, Theoharides TC. The effects of plant flavonoids on mammalian cells: Implications for inflammation, heart disease, and cancer. Pharmacol Rev 2000;52:673–751. 17. Beecher GR. Overview of dietary flavonoids: Nomenclature, occurrence and intake. J Nutr 2003;133:3248S–3254S. 18. Guajardo-Flores D, Serna-Saldivar SO, Gutierrez-Uribe JA. Evaluation of the antioxidant and antiproliferative activities of extracted saponins and flavonols from germinated black beans (Phaseolus vulgaris L.). Food Chem 2013;141:1497–1503. 19. Sahu BD, Kalvala AK, Koneru M, et al. Ameliorative effect of fisetin on cisplatin-induced nephrotoxicity in rats via modulation of NF-kappaB activation and antioxidant defence. PLoS ONE 2014;9:e105070. 20. Pal HC, Sharma S, Elmets CA, Athar M, Afaq F. Fisetin inhibits growth, induces G2 /M arrest and apoptosis of human epidermoid carcinoma A431 cells: Role of mitochondrial membrane potential disruption and consequent caspases activation. Exp Dermatol 2013;22:470–475. 21. Lu X, Jung J, Cho HJ, et al. Fisetin inhibits the activities of cyclin-dependent kinases leading to cell cycle arrest in HT-29 human colon cancer cells. J Nutr 2005;135:2884–2890. 22. Ravichandran N, Suresh G, Ramesh B, Manikandan R, Choi YW, Vijaiyan Siva G. Fisetin modulates mitochondrial enzymes and apoptotic signals in benzo(a)pyrene-induced lung cancer. Mol Cell Biochem 2014;390:225–234. 23. Li J, Cheng Y, Qu W, et al. Fisetin, a dietary flavonoid, induces cell cycle arrest and apoptosis through activation of p53 and inhibition of NF-kappa B pathways in bladder cancer cells. Basic Clin Pharmacol Toxicol 2011;108:84–93. 24. Khan N, Afaq F, Syed DN, Mukhtar H. Fisetin, a novel dietary flavonoid, causes apoptosis and cell cycle arrest in human prostate cancer LNCaP cells. Carcinogenesis 2008;29:1049–1056. 25. Bhat TA, Nambiar D, Pal A, Agarwal R, Singh RP. Fisetin inhibits various attributes of angiogenesis in vitro and in vivoimplications for angioprevention. Carcinogenesis 2012;33: 385–393. 26. Mateen S, Tyagi A, Agarwal C, Singh RP, Agarwal R. Silibinin inhibits human nonsmall cell lung cancer cell growth through cell-cycle arrest by modulating expression and function of key cell-cycle regulators. Mol Carcinog 2010;49:247–258. 27. Yim D, Singh RP, Agarwal C, Lee S, Chi H, Agarwal R. A novel anticancer agent, decursin, induces G1 arrest and apoptosis in human prostate carcinoma cells. Cancer Res 2005;65: 1035–1044. 28. Varghese L, Agarwal C, Tyagi A, Singh RP, Agarwal R. Silibinin efficacy against human hepatocellular carcinoma. Clin Cancer Res 2005;11:8441–8448. 29. Tice RR, Agurell E, Anderson D, et al. Single cell gel/comet assay: Guidelines for in vitro and in vivo genetic toxicology testing. Environ Mol Mutagen 2000;35:206–221.

16

SABARWAL ET AL.

30. Roy S, Kaur M, Agarwal C, Tecklenburg M, Sclafani RA, Agarwal R. P21 and p27 induction by silibinin is essential for its cell cycle arrest effect in prostate carcinoma cells. Mol Cancer Ther. 2007;6:2696–2707. 31. Singh T, Sharma SD, Katiyar SK. Grape proanthocyanidins induce apoptosis by loss of mitochondrial membrane potential of human non-small cell lung cancer cells in vitro and in vivo. PLoS ONE 2011;6:e27444. 32. De Paepe B, Smet J, Vanlander A, et al. Fluorescence imaging of mitochondria in cultured skin fibroblasts: A useful method for the detection of oxidative phosphorylation defects. Pediatr Res 2012;72:232–240. 33. Kuo LJ, Yang LX. Gamma-H2AX—A novel biomarker for DNA double-strand breaks. In Vivo (Athens, Greece) 2008;22: 305–309. 34. Ambrosini G, Adida C, Altieri DC. A novel anti-apoptosis gene, survivin, expressed in cancer and lymphoma. Nat Med 1997;3:917–921. 35. Tsuburaya A, Noguchi Y, Yoshikawa T, et al. An antiapoptosis gene, survivin and telomerase expression in gastric cancer. Hepatogastroenterology 2002;49:1150–1152. 36. Smith ML, Seo YR. P53 regulation of DNA excision repair pathways. Mutagenesis 2002;17:149–156. 37. Dewaele M, Maes H, Agostinis P. ROS-mediated mechanisms of autophagy stimulation and their relevance in cancer therapy. Autophagy 2010;6:838–854. 38. Sesti F, Tsitsilonis OE, Kotsinas A, Trougakos IP. Oxidative stress-mediated biomolecular damage and inflammation in tumorigenesis. In Vivo (Athens, Greece) 2012;26:395–402. 39. Ricci JE, Waterhouse N, Green DR. Mitochondrial functions during cell death, a complex (I-V) dilemma. Cell Death Differ 2003;10:488–492.


40. Murphy MP. How mitochondria produce reactive oxygen species. Biochem J 2009;417:1–13. 41. Starkov AA. The role of mitochondria in reactive oxygen species metabolism and signaling. Ann NY Acad Sci 2008;1147:37–52. 42. Abbas T, Dutta A. P21 in cancer: Intricate networks and multiple activities. Nat Rev Cancer 2009;9:400–414. 43. Ekholm SV, Reed SI. Regulation of G(1) cyclin-dependent kinases in the mammalian cell cycle. Curr Opin Cell Biol 2000;12:676–684. 44. Shi Y. A structural view of mitochondria-mediated apoptosis. Nat Struct Biol 2001;8:394–401. 45. Jang KY, Jeong SJ, Kim SH, et al. Activation of reactive oxygen species/AMP activated protein kinase signaling mediates fisetin-induced apoptosis in multiple myeloma U266 cells. Cancer Lett 2012;319:197–202. 46. Maurya BK, Trigun SK. Fisetin modulates antioxidant enzymes and inflammatory factors to inhibit aflatoxin-B1 induced hepatocellular carcinoma in rats. Oxid Med Cell Longevity 2016;2016:1–9. 47. Kang KA, Piao MJ, Kim KC, et al. Fisetin attenuates hydrogen peroxide-induced cell damage by scavenging reactive oxygen species and activating protective functions of cellular glutathione system. In Vitro Cell Dev Biol Anim 2014;50: 66–74.

SUPPORTING INFORMATION Additional supporting information may be found in the online version of this article at the publisher’s web-site.