Induction of sestrin 2 is associated with fisetin-mediated apoptosis in human head and neck cancer cell lines

Fisetin was reported to have an anti-proliferative and apoptotic activity as a novel anti-cancer agent in various cancer cell lines. However, the possible molecular targets for the anti-cancer effect of fisetin in human head and neck cancer (HNCC) have not yet been clarified. In this study, the influence of fisetin on the growth and apoptosis of HNCCs were examined. In HSC3 cells, fisetin treatment reduced the viability and induced apoptosis. Through the results from the screening of the expression profile of apoptosis-related genes, sestrin 2 (SESN2) was functionally involved in fisetin-mediated apoptosis showing the knockdown of SESN2 by siRNA clearly restored fisetin-induced apoptosis. In addition, fisetin reduced the protein expression levels of phospho-mTOR (p-mTOR) and Mcl-1, which are the downstream molecules of SESN2. It also induced PARP cleavage by inducing an increase in the expression levels of SESN2 together with reducing mTOR and Mcl-1 proteins in other three HNCCs (MC3, Ca9.22, and HN22). Taken together, our findings suggest that the anti-cancer effect of fisetin on HNCCs is associated with SESN2/mTOR/Mcl-1 signaling axis.

Introduction S estrins (SESNs), a protein family composed of SESN1, SESN2, and SESN3 in mammals, are conserved stressresponsive proteins that reduce reactive oxygen species (ROS) and are involved in the regulation of cell survival. (1) SESN2, a homolog of p53-activated gene 26, is induced by cytotoxic events such as hypoxia, DNA damage, and oxidative stress. (2,3) SESN2 is down-regulated in cancer cells as a result of the increased production and accumulation of ROS and the redox state of tumor cells. (4,5) Similarly, a lack of SESN2 in mouse embryonic fibroblasts increases proliferation of RAS-activated tumor cells. (6) Recently, several studies have reported that SESN2 suppressed cell proliferation and was involved in apoptosis in colon cancer cells. (1,7) However, the relationship between SESN2 and oral cancer is not clear.
Fisetin (3,3',4',7-tetrahydroxyflavone) is a naturally occurring flavonoid found in numerous vegetables and fruits such as apple, onion, grape, cucumber, persimmon, and strawberry. (8,9) Antidiabetic, cardio-protective, and neuro-protective activities of fisetin have been demonstrated by using in vitro and in vivo experimental models relevant to human diseases. (10)(11)(12) A potential against cell growth and survival of various cancer cells has been shown. (13)(14)(15) Recently, fisetin inhibited malignant proliferation in human oral squamous cell carcinoma cell lines through inhibition of Met/Src signaling pathways. (16) However, crucial molecular targets for the anticancer effect of fisetin have not been identified on human head and neck cancer cells (HNCCs).
Here, the anticancer activity and the molecular targets of fisetin in HNCCs were investigated in vitro. Our results indicate that fisetin induces apoptosis in HNCCs by upregulating SESN2.

Materials and Methods
Cell culture and chemical treatment. HSC3 and Ca9. 22 Human oral squamous cell carcinoma cell lines were provided from Prof. Shindoh (Hokkaido University, Sapporo, Japan). MC3 mucoepidermoid carcinoma cell line was provided by prof. Wu Junzheng (Forth Military Medical University, Xi'an, China), and HN22 human head and neck squamous cell carcinoma cell line was obtained from Prof. Lee (Dankook University, Cheonan, Korea). The cells were cultured in DMEM supplemented with 10% fetal bovine serum (FBS) and antibiotics at 37°C in 5% CO 2 incubator. All experiments were prepared after the cells reached 50~60% confluence. Fisetin ( Fig. 1A; Sigma-Aldrich, St. Louis, MO) was dissolved in dimethyl sulfoxide (DMSO), aliquoted, and stored at -20°C. Final concentration of DMSO did not exceed 0.1%.
Trypan blue exclusion assay. The growth inhibitory effects of fisetin were determined with trypan blue solution (Gibco, Paisley, UK). Cells were stained with trypan blue (0.4%), and viable cells were counted using a hemocytometer.
Western blotting. Whole-cell lysates were prepared with a lysis buffer and protein concentration of each sample was measured using a DC Protein Assay Kit (BIO-RAD Laboratories, Madison, WI). After normalization, equal amount of protein was separated by SDS-PAGE and transferred to Immuno-Blot PVDF membranes. The membranes were blocked with 5% skim milk in TBST at RT for 2 h and incubated with primary antibodies and corresponding HRP-conjugated secondary antibodies. Antibodies against cleaved PARP, cleaved caspase-3, SESN2, p-mTOR, mTOR, and Mcl-1 were purchased from Cell Signaling Technology, Inc. (Charlottesville, VA) and actin antibody was obtained from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). The immunoreactive bands were visualized by ImageQuant LAS 500 (GE Healthcare Life Sciences, Piscataway, NJ).
Live/dead assay. The cytotoxicity of fisetin was determined using a Live/Dead Viability/Cytotoxicity assay kit (Life Tech-S nologies, Grand Island, NY). The polyanionic dye, calcein-AM is retained within live cells, producing an intense green fluorescence through intracellular esterase activity. Ethidium homodimer-1 enters dead cells with damaged cell membranes and binds to nucleic acids, producing a bright red fluorescence. Briefly, cells were stained with 2 mM calcein-AM and 4 mM ethidium homodimer-1 and incubated for 30 min at RT. Cells were analyzed under a fluorescence microscopy (Leica DMi8, Wetzlar, Germany) with the appropriate excitation and emission filters. A total of three random photo were selected from each three independent experiments for quantification. The percentage of live cells was manually calculated by measuring the number of green fluorescence-labeled cells.
Microarray. Total RNA was extracted from cells using RNeasy Mini kit (Qiagen, CA) according to the manufacturer's instructions. Two sets of samples were independently prepared and analyzed. The integrity and quantity of total RNA were assessed by Agilent 2100 Bioanalyzer and Nanodrop 1000 analyzer. For each sample, total RNA was analyzed using a Human Gene 2.0 ST Array. The GeneChip Arrays were immediately scanned with Affymetrix GeneChip Scanner 3000 7 G.
Real Time PCR. Total RNA was extracted using easy-BLUE Total RNA Extraction Kit (INTRON, Daejeon, Korea). The isolated RNA was transcribed by AMPIGENE cDNA Synthesis Kit (Enzo Life sciences, Inc., NY) and real time PCR was performed using the StepOne Real-Time PCR System with AMPI-GENE qPCR Green Mix Hi-Rox (Enzo Life sciences, Inc., Farmingdale, NY). Real-time PCR conditions for all genes were as follows: 95°C for 2 min, followed by 40 cycles of 95°C for 10 s and 62°C for 30 s. The relative expression changes of the target genes were quantified by normalizing their expression to that of GAPDH. The PCR primers of all the target genes are listed in Table 1.
siRNA transfection. siSESN2 and siControl were purchased from Bioneer (Daejeon, Korea). Briefly, HSC3 cells were transfected with 50 nM siRNA using Lipofectamine 3000 (Invitrogen, Carlsbad, CA) for 6 h followed by treatment with fisetin for 24 h. After the transfection, HSC3 cells were subjected to Western blotting.
Statistical analysis. ANOVA followed by Tukey's post hoc test were used to determine the significance between the control and treatment groups; p value of <0.05 was considered significant.

Results
Effects of fisetin on growth and apoptosis of HSC3 human head and neck cancer cells. To determine the anti-proliferative activity of fisetin on HSC3 cells, cells were treated with fisetin at various concentrations (7.5, 15, and 30 mM) for 24 h and the cell viability was measured using a trypan blue exclusion assay. Fisetin inhibited the viability of HSC3 cells in a concentrationdependent manner Fig. 1B). The association of the apoptosis and the growth inhibitory effects of fisetin was assessed by western blotting with an antibody capable of detecting cleaved PARP. The PARP cleavage was significantly increased in fisetin-treated groups (Fig. 1C). The fisetin-treated HSC3 cells were also stained with the live/dead assay kit or DAPI solution. The dead cells (red stained) were significantly increased after the fisetin treatment ( Fig. 1D). In addition, the fisetin-treated HSC3 cells showed prominent nuclear condensation and fragmentation (Fig. 1E).
These results indicate that fisetin inhibits cell growth and induces apoptosis in human head and neck cancer cells (HNCCs).

Involvement of SESN2 in fisetin induced apoptosis in
HSC3 cells. To identify the possible molecular targets for fisetin-induced apoptosis, HSC3 cells were incubated in the presence or absence of 30 mM fisetin and cDNA microarray analysis was performed using Human Gene 2.0 ST Array. The calculated upregulation or downregulation of gene expressions, as determined by the microarray, are listed in Table 2 and 3, respectively. As shown in Table 2 and 3, fisetin-treated HSC3 cells up-regulated 61 genes over 2 folds and down-regulated 81 genes over 2 folds compared with DMSO-treated cells. To verify the reliability, real-time PCR was performed with primers of three different genes, each of which increased (C1orf162, ARRDC4, and FCER1G) or decreased (MANSC1, EPHX4, and GBP3). The results from real-time PCR were similar to those of microarray ( Fig. 2B) suggesting that microarray data are reliable. Among 142 genes, 7 apoptosis-associated genes were selected and SESN2 and CHAC1 (ChaC, cation transport regulator homolog 1) changed overwhelmingly compared with other 5 genes (Fig. 2C). Next, we evaluated whether fisetin affects the expression levels of both SESN2 and CHAC1 proteins in HSC3 cells and we observed a significant increase in SESN2 protein expression in HSC3 cells treated with fisetin ( Fig. 3A) while CHAC 1 protein was not affected (data not shown). To investigate the functional consequence of the increased SESN2, HSC3 cells were transfected with either siSESN2 or siCon. As shown in Fig. 3B, the knockdown of SESN2 by siSESN2 resulted in significantly less fisetin-induced apoptosis compared to the cells transfected with siCon. These data indicated that SESN2 mediates the apoptosis induced by fisetin in HSC3 cells. Since SESN2 is known to regulate mammalian target of rapamycin (mTOR) and myeloid cell leukemia-1 (Mcl-1) during apoptosis, (1,17) the effects of fisetin on mTOR/Mcl-1 was determined. The results showed that fisetin reduced the protein expression levels of p-mTOR and Mcl-1 (Fig. 3C). Taken together, these findings suggest that fisetininduced apoptosis may be related to SESN2/mTOR/Mcl-1 signaling axis.     (Fig. 4B). These findings suggest that fisetin-induced apoptosis via SESN2 in HNCCs is a general mechanism for the anticancer effect of fisetin.

Discussion
Head and neck cancer (HNCC) is diagnosed in more than 63,000 people in the United States and is responsible for approximately 13,300 death annually in United State. (18) HNCC is a challenging clinical problem and it is recognized that there is a need to develop alternative methods for the management of this tumor. Naturally occurring compounds from plants, vegetables, and fruits have long been used in traditional medicinal systems because of their non-toxic nature in effective dosages. (20) Thus, there is an increasing interest in the efficacy analysis of inhibiting the proliferation of HNCC using phytochemicals with no toxicity. Our research group recently found naturally-derived chemicals such as silymarin, caffeic acid phenethyl ester, and oridonin have anticancer efficacies by inducing apoptosis of HNCCs. (21)(22)(23) Fisetin is also naturally derived compound that can be easily synthesized and has no toxicity to normal cells. (24) First, we demonstrated that fisetin reduced cell viability and induced apoptosis in four different HNCCs (HSC3, MC3, Ca9.22, and HN22). There are several evidences to support our present data showing that fisetin suppresses malignant proliferation and induces apoptosis in human oral squamous cell carcinoma and laryngeal carcinoma. (16,24) As mentioned earlier, fisetin has been known to have anticancer efficacies in various types of cancers. It inhibited cell proliferation of cancer cells through modulation of multiple signaling pathways. Fisetin-induced apoptosis of human oral cancer cells was through ROS production and mitochondria-dependent signaling pathways. (25) Fisetin also induced apoptosis through p53-mediated upregulation of death receptor 5 expression in human renal carcinoma cells. (26) However, the molecular targets for fisetininduced apoptosis in HNCCs are not yet known precisely. Here, using gene expression profiling, we found that fisetin treatment affected 142 genes (61 up-regulated genes and 81 down-regulated genes). Among them, the mRNA expression levels of SESN2 and CHAC1, which are apoptosis-associated genes, were significantly up-regulated in fisetin-treated HSC3 cells. Fisetin also increased the expression level of SESN2 protein and knockdown of SESN2 by a siRNA technique clearly recovered HSC3 cells from fisetininduced apoptosis, suggesting that SESN2 is involved in fisetinmediated apoptosis. Consistent with our data, it was reported that quercetin, which flavonoid and structurally similar to fisetin, regulates the SENS2 signaling and induces apoptosis by inducing the generation of intracellular ROS and AMPK/p38 signaling pathway. (27) It suggests the possibility of SESN2 as a therapeutic target for cancer treatment. To our best knowledge, this is the first report demonstrating that SESN2 may be a molecular target for fisetin-induced apoptosis in HNCCs. CHAC1 was identified as a component of the unfolded protein response (UPR) signaling pathway. Fisetin induced apoptosis through endoplasmic reticulum (ER) stress in oral cancer and non-small cell lung cancer. (25,28,29) However, we found that fisetin treatment did not alter the expression level of CHAC1 protein while it significantly augmented its mRNA level in the present study. These data indicate that fisetininduced apoptosis may not be associated with ER stress in HSC3 HNCCs.
The mammalian target of rapamycin (mTOR) is a serinethreonine protein kinase that are resistant to the growth-inhibitory activities of rapamycin. mTOR comprises of two distinct multiprotein complexes; mTORC1 and mTORC2. SESN2 inhibits mTOR-dependent phosphorylation of p70S6K and 4E-BP1, (6) and knockdown of SESN2 resulted in the activation of mTOR signaling indicating that the important role of SESN2 on mTOR inhibition. (30) Our results from western blotting demonstrated that mTOR was down-regulated by fisetin treatment in HNCCs consistent with other previous studies showing tumor inhibiting potentials of fisetin through mTOR pathway. (24,31,32) These results suggest that fisetin-induced apoptotic potentials is through SESN2/ mTOR signaling axis.
In conclusion, fisetin suppresses the growth of HNCCs and induces apoptotic cell death. Its functional role is attributed to SESN2/mTOR/Mcl-1 signaling axis. On the basis of our findings, we suggest that targeting SESN2 by fisetin or its derivatives could be a new strategy to overcome HNCCs.

Author Contributions
DH Won and SH Chung performed experiments and drafted the manuscript; JA Shin, KO Hong, IH Yang were responsible for the study design and data analysis; JW Yun and SD Cho designed the study and revised the manuscript. All authors reviewed and approved the final manuscript.