The human GC cell line, MGC803, was obtained from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China). The cells were cultured in RPMI-1640 medium (Gibco-BRL, Grand Island, NY, USA) containing 10% fetal bovine serum (FBS) (Gibco, Rockville, MD, USA), 100 kU/l penicillin, 0.1 g/l streptomycin, 0.3 g/l L-glutamine (Invitrogen, Carlsbad, CA, USA) and 0.85 g/l NaHCO₃ at 37°C in a humidified atmosphere containing 5% CO₂. Spheroid bodies were derived from parental cells. Briefly, the parental cells were first cultured in serum-free RPMI-1640 medium with 20 ng/ml human fibroblast growth factor 2 (FGF-2), 100 ng/ml epidermal growth factor (EGF) (Chemicon, Japan), 1% N-2 supplement, 2% B-27 supplement (Invitrogen) and 1% antibiotic mixture (Gibco). First, parental GC cells were plated into 96-well ultra-low attachment plates at 1,000 cells/well. Subsequently, the formation of spheroid bodies was observed under an inverted microscope once in a while. Primary spheroid bodies were noted within 2 weeks. The spheroid bodies were then collected and dissociated by trypsinization for passage and amplification in serum-free medium.

The subspheroid body formation assay was performed using crystal violet staining. The collected spheroid bodies were dissociated and centrifuged into single suspended cells. The cells were seeded into 6-well plates at a density of 1×10² cells/well, as well as the digested GC cells. The serum-free culture medium was changed every 3–4 days. Two weeks later, the culture medium was removed and the cells were washed in PBS 3 times. Subsequently, 4% paraformaldehyde was added for 30 min to fix the cells, which were then stained with crystal violet (Sigma, St. Louis, MO, USA) for 10 min. Obvious cell colonies were observed under an inverted microscope following 10 washes with PBS. Colonies containing <50 cells were excluded.

A Qiagen RNeasy mini kit (Qiagen, Valencia, CA, USA) was used for total RNA extraction from the parental cells and spheroid body-forming cells. Agarose gel electrophoresis was used to detect the integrity and purification of RNA samples. Absorption under 260 nm was used to determine the RNA concentration. cDNA synthesis was conducted using a High-Capacity cDNA Reverse Transcription kit (Applied Biosystems, Foster City, CA, USA); qPCR was performed with SYBR-Green (Takara, Dalian, China) as a fluorescent dye to determine the changes in the levels of stem-specific genes [octamer-binding transcription factor 4 (Oct4), SRY (sex determining region Y)-box 2 (Sox2), Nanog and CD44] following the manufacturer's instructions. β-actin was used as a normalizing control. The primer sequences were as follows: Oct4, 5′-AGCCCTCATTTCACCAGGCC-3′ (forward) and 5′-CCCCCACAGAACTCATACGG-3′ (reverse); Sox2, 5′-GAAAAACGAGGGAAATGGG-3′ (forward) and 5′-GCTGTCATTTGCTGTGGGT-3′ (reverse); Nanog, 5′-CTCCTCCCATCCCTCATA-3′ (forward) and 5′-AGGCTCCAACCATACTCC-3′ (reverse); CD44, 5′-CATCCCAGACGAAGACAGTCC-3′ (forward) and 5′-TGATCAGCCATTCTGGAATTTG-3′ (reverse); and β-actin, 5′-CTGAGAGGGAAATCGTGCGT-3′ (forward) and 5′-CCACAGGATTCCATACCCAAGA-3′ (reverse).

An in vitro analysis of the cytotoxic effects of quercetin on GCSCs was carried out by MTT assay. In brief, 200 µl cells were seeded into a 96-well plate and incubated for 12 h, and the medium was then removed. Following treatment with various concentrations of quercetin (0, 20, 40, 60, 80 and 100 µM) for the indicated perdios of time at 37°C, 20 µl 5 mg/ml MTT (Sigma) solution were added to each well. Following a 4 h of incubation, DMSO (100 µl) was added to dissolve the formazan crystals for 15 min. The optical density at 570 nm was measured to calculate the inhibitory function of quercetin from the cytotoxicity curves. Cells without treatment were used as controls. Quercetin (99% pure) was purchased from Sigma, dissolved in DMSO, and stored at −20°C. Its structure is illustrated in Fig. 1A. The half maximal inhibitory concentration (IC₅₀) was defined as the drug concentration which inhibited cell growth by 50%, and was calculated using a linear regression based on the results of MTT assay.

The cells were counted and transferred to a 5-ml tube, washed with HBSS containing 2% heat-inactivated FBS, and we reconstituted cells at a concentration of 10⁶/100 µl with Sandoglobulin solution (Sandoz Pharmaceuticals Corp., East Hanover, NJ, USA). Subsequently, the cells were incubated on ice for 20 min. After washing twice with HBSS, the cells were resuspended in HBSS containing 1 mM HEPES and 2% FBS. After filtering with a sterile mesh filter (40 µm), the cells were incubated with anti-CD44-FITC at a dilution of 1:400 (no. 560977; BD Biosciences, Mountain View, CA, USA) for 30 min. The cells were then rinsed with HBSS and subsequently resuspended in HBSS solution with 2% FBS, 1 mM HEPES and 1% penicillin-streptomycin sulfate. Subsequently, the samples were examined using FACSAria (serial no. 342975; BD Immunocytometry Systems, Franklin Lakes, NJ, USA) for fluorescence-activated cell sorting. The cells were routinely sorted twice, as previously described (26).

The quantification of the apoptosis of the GCSCs treated with quercetin was examined by flow cytometry. The cells were treated with 53.0 µM quercetin for 24 h, and then centrifuged at 1,000 × g for 5 min, and washed twice with PBS. An Annexin V-fluorescein isothiocyanate (FITC) apoptosis detection kit (Invitrogen) was used for the detection of apoptotic cells. After staining with 5 µl Annexin V-FITC and 5 µl propidium iodide for 15 min in the dark at room temperature, the cells were immediately subjected to flow cytometric observation and analysis within 1 h.

The resultant fluorescence intensity was used as an arbitrary unit representing mitochondrial membrane potential. The GCSCs were treated with the AKT activator, fumonisin B1 (FB1, 25 µM) (Sigma) for 2 h, in prior to exposure to quercetin for 24 h and rinsed with PBS. The cells were resuspended in 1 ml rhodamine 123 (1 µg/ml; Sigma) for 30 min in the dark at 37°C. After rinsing with PBS 3 times, the fluorescence intensity of the cells was examined by fluorescence-activated cell sorting at 507 nm (excitation) and 529 nm (emission).

Caspase activity was measured using a caspase colorimetric assay kit from Sigma. Briefly, the cells were treated with quercetin for 24 h and then lysed. The centrifugal supernatant was placed into a 96-well microplate with reaction buffer. The microplate was incubated at 37°C overnight and substrate was added. The activity was measured at 405 nm using a spectrophotometer (U-5100; Techcomp Ltd., Shanghai, China).

Before western blot analysis was performed, the cells were lysed with RIPA lysis buffer [100 mM NaCl, 50 mM Tris-HCl pH 7.5, 1% Triton X-100, 1 mM ethylenediaminetetraacetic acid (EDTA), 10 mM β-glycerophosphate and 2 mM sodium vanadate and protease inhibitor] to collect whole-cell extracts. To prepare the cytosolic extracts, the cells were lysed with RIPA lysis buffer containing protease inhibitor cocktail for 30 min on ice, and were then homogenized by 6 passages through a 25-gauge needle. Following centrifugation (14,000 × g) at 4°C for 10 min, the resultant supernatant (cytosol) was subjected to western blot analysis to evaluate the release of cytochrome c (Cyt-c). The protein samples were then separated using SDS-PAGE on a 12% gel and transferred onto polyvinylidene difluoride membranes (Amersham Pharmacia, Freiburg, Germany). Non-specific binding was blocked by incubation with 5% non-fat milk in Tris-buffered saline with Tween-20 (TBST) buffer at room temperature for 1 h. Incubation with the primary antibody was carried out overnight in a cold room, and the membranes were subsequently washed in TBST 3 times. Primary antibodies were as follows: anti-Bax (no. 2774, 1:400), anti-Bcl-2 (no. 2872, 1:200), anti-Cyt-c (no. 4272, 1:300) and anti-β-actin (no. 4967S) were purchased from Cell Signaling Technology (Danvers, MA, USA). A secondary antibody conjugated to horseradish peroxidase (1:5,000; Sigma) was added to the membranes followed by incubation for 1 h at room temperature, followed by enhanced chemiluminescence using an ECL kit (Amersham Pharmacia, Piscataway, NJ, USA). Finally, protein bands were visualized using ECL and quantified using Quantity One software (Bio-Rad, Hercules, CA, USA).

All experimental data were derived from at least 3 independent experiments, and all values are expressed as the means ± SD. Statistical analyses were performed using the SPSS 16.0 statistical software package (SPSS Inc., Chicago, IL, USA). Any differences were assessed using the Student's t-test or one-way ANOVA. A value of p<0.05 was considered to indicate a statistically significant difference.

The MGC803 parental cells cultured in serum-free medium grew as non-adherent, three-dimensional spheroid clusters, termed spheroid bodies. As time progressed, the cell spheroid bodies gradually increased not only in volume, but also in number. The self-renewal capacity of these spheroid body-forming cells was preliminarily assessed by counting the subspheroid bodies that they generated. As shown in Fig. 1B, compared with the parental cells, the rate of spheroid body formation was 25.4±3.51% in the spheroid body cells, whereas for the parental cells it was only 5.0±1.04%, reflecting the high capacity for self-renewal of these isolated spheroid body cells. Furthermore, the results of RT-qPCR revealed that the mRNA levels of stem-specific Oct4, Sox2, Nanog and CD44 in the spheroid body-forming cells were significantly higher than those in the parental cells (Fig. 1C). In addition, the results of western blot analysis revealed the high expression of the stem-cell markers, Oct4, Sox2, Nanog and CD44, in the spheroid body cells, as compared to the parental cells (Fig. 1D). The relative protein levels of CD44, Nanog, Sox2 and Oct4 in the spheroid body-forming cells were approximately double those of the parental cells, at least for CD44, Nanog and Sox2 (Fig. 1E). Taken together, these results indicate the successful isolation of GCSCs from the MGC803 GC cells.

To determine the cytotoxity of quercetin on the GCSCs, we used an MTT assay to examine cell viability. As shown in Fig. 2A, treatment with quercetin significantly increased the inhibition rate of viable cells in a time- and dose-dependent manner. Moreover, after treating the GCSC cells with 20 and 100 µM quercetin for 24 h, cell viability decreased to 82±6.73% (20 µM) and 20.5±2.10% (100 µM) (Fig. 2B), and quercetin strongly restrained GCSC cell viability with an IC₅₀ of approximately 53 µM. Additionally, we noted that the MGC803 cells were more sensitive than the GCSCs to quercetin, reflecting the relative stronger chemoresistance of GCSCs in comparison to the parental cells. Moreover, a decrease of >50% in the number of Oct4⁺CD44⁺ cells was observed following treatment with quercetin (Fig. 2C), indicating the diminished proliferative ability of GCSCs upon treatment with quercetin.

In order to further analyze the effects of quercetin on GCSC survival, flow cytometry was used to assess quercetin-induced apoptosis. As shown in Fig. 2D, in comparison with the control (untreated) group, quercetin significantly induced apoptosis. When the cells were treated with 53.0 µM quercetin for 24 h, GCSC apoptosis was significantly enhanced (Fig. 2E). However, the sensitivity of the GCSCs to quercetin was less than that of the MGC803 cells. These data reflect the inhibitory effects which quercetin exerts on the growth of GCSCs; however, the mechanisms involved require further investigation.

Mitochondrial apoptosis plays an important role in GC growth inhibition. To ascertain the underlying mechanisms involved in the quercetin-triggered inhibition of cell growth, the mitochondrial apoptotic pathway was analyzed. As shown in Fig. 3A, quercetin decreased mitochondrial membrane potential in a time-dependent manner. The fluorescence intensity at 12 h was higher than after 24 h of treatment with quercetin. Concomitantly, as central molecules of the apoptotic cascade, caspase-3 and caspase-9 were also investigated, and we noted that quercetin upregulated the expression levels of both caspase-3 and caspase-9 (Fig. 3B). Our findings demonstrated that quercetin-induced apoptotic cell death was associated with the activation of caspase-3 and caspase-9. In order to further explore the involvement of the mitochondrial apoptotic pathway in the quercetin-induced inhibition of GCSC growth, the expression of Bax, Bcl-2 and Cyt-c was examined by western blot anlaysis. As shown in Fig. 3C, an increase in the levels of Bax and Cyt-c occurred, together with a significant decrease in the expression of the anti-apoptotic protein, Bcl-2, upon treatment with quercetin. The variations in the levels of these proteins can be directly noted in Fig. 3D. These results suggest that quercetin inhibits the growth of GCSCs by regulating the mitochondrial apoptotic pathway.

Previous studies have demonstrated that quercetin induces cell apoptosis in many types of cancer by blocking the phosphoinositide 3-kinase (PI3K)-Akt pathway, which plays a negative role in the intrinsic mitochondrial pathway (27–29). Therefore, in the present study, the phosphorylation of Akt was investigated. As was expected, an attenuated level of p-Akt was observed following treatment with quercetin (Fig. 4A). Treatment with quercetin induced only a slight decrease in the expression of Akt (Fig. 4B). On the other hand, the addition of fumonisin B1 (FB1, Akt activator) significantly attenuated the inhibitory effects of quercetin on Akt phosphorylation (Fig. 4A). Moreover, the quercetin-induced inhibition of GCSC viability was almost completely abrogated once Akt was activated by treatment with FB1 (Fig. 4C). At the same time, the upregulation of cell apoptosis induced by quercetin was also markedly decreased when the cells were pre-treated with FB1 (Fig. 4D). These results suggest that the PI3K-Akt signaling pathway is involved in the quercetin-induced inhibition of the growth of GCSCs.

In order to gain further insight into the involvement of Akt in the mitochondrial apoptotic pathway, we examined the influence of FB1 on the quercetin-induced mitochondrial pathway. As shown in Fig. 5A, treatment with quercetin significantly reduced mitochondrial membrane potential, while the addition of FB1 reversed this effect. FB1 markedly abated the increase in the level of caspase-3 (Fig. 5B) and caspase-9 (Fig. 5C) induced by quercetin. Furthermore, the effects of quercetin on the expression of Cyt-c, Bax and Bcl-2 were attenuated by the Akt activator, FB1 (Fig. 5D). Their relative levels were calculated and are depicted in Fig. 5E. These results clearly illustrate the involvement of Akt in the quercetin-induced activation of the mitochondrial apoptotic pathway.