N-acetyl-L-cysteine (NAC), apigenin, 2′-7′-dichlorodihydrofluorescein diacetate (DCF-DA) and compound 968 were purchased from Sigma-Aldrich (St. Louis, MO, USA) or Millipore (Billerica, MA, USA). Antibodies recognizing cleaved caspase-9, cleaved caspase-3, PARP, and HK1 (catalog nos. 9502, 9661, 9546, and 2024, respectively) were purchased from Cell Signaling Technology (Danvers, MA, USA). GLUT1 (sc-7903), GLS1 (156876) and β-tubulin (sc-9104) were purchased from Santa Cruz Biotechnology (Dallas, TX, USA) and Abcam (Cambridge, MA, USA).

Lung cancer cells were cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum and 25 mM glucose. The lentiviral vector pLenti-GIII-CMV- SLC2A1, expressing human GLUT1 was obtained from Applied Biological Materials Inc. (Richmond, BC, Canada). A short hairpin RNA construct targeting GLS1 was purchased from Sigma-Aldrich. The hairpin sequence was: 5′-CCGGG CCCTGAAGCAGTTCGAAATACTCGAGTATTTCGAAC TGCTTCAGGGCTTTTTG-3′. Lentiviral particles were produced in HEK293T cells, harvested 48 h after transfection, and then used to infect lung cancer cells. Infected lung cancer cells were selected by growth on media containing 2 μg/ml of puromycin (Invitrogen, Carlsbad, CA, USA) for 6 days.

Cells were lysed in buffer containing 1% IGEPAL, 150 mM NaCl, 50 mM Tris-HCl (pH 7.9), 10 mM NaF, 0.1 mM EDTA, and protease inhibitor cocktail. Protein samples were separated by SDS-PAGE, and the separated proteins were then transferred to a PVDF membrane (Millipore). Membranes were incubated in the presence of primary antibodies (1:1,000) at 4°C overnight, and HRP-conjugated secondary antibodies (1:10,000) were incubated for 1 h at room temperature. Proteins were visualized using an enhanced chemiluminescence (ECL) Prime kit (GE Healthcare, Pittsburgh, PA, USA).

To measure ROS levels, cells were incubated with Hank's balanced salt solution (HBSS), containing 30 μM DCF-DA, for 30 min. After incubation, cells were detached and resuspended in phosphate-buffered saline (PBS) in preparation for fluorescence-activated cell sorting (FACS) analysis.

Total RNA was extracted from cells using TRIzol (Invitrogen), and 2 μg of total RNA was used for cDNA synthesis using a high capacity cDNA reverse transcription kit (Applied Biosystems). Quantitative PCR was performed using SYBR Green PCR Master Mix (Applied Biosystems). The sequences of the PCR primers (5′-3′) were: GAGAGTTTCATCGGAGAGCC and CAGCGAGAATCGG CTACAG for HK1; GGCATTGATGACTCCAGTGTT and ATGGAGCCCAGCAGCAA for GLUT1; CTTGGGACAG CAGCCTTAAT and CAAGCTGGACGTTAAAGGGA for PGK1; AGCCCTTTCTCCATCTCCTT and AACCATGAC CAAGTGCAGAA for HK2; TACAGGCACAGTCGCAG AGT and CACTTCCTGGATGCTTGCTG for ALDOA; CCTGGCATGGATCTTGAGAA and TACGTTCACCT CGGTGTCTG for ENO1.

As previously described (9), lactate and glucose levels in cultured media were measured using lactate colorimetric assay kit and glucose colorimetric assay kit, respectively (BioVision, USA). Briefly, cells were cultured in DMEM without phenol red for 24 h, and then cultured media were mixed with assay solution. To measure intracellular ATP levels, cells were cultured in 6-well plates, and cell lysates were mixed with reaction buffer. ATP levels in the cell lysates were then determined using a luciferin-luciferase-based assay kit (Invitrogen), used according to the manufacturer's instructions. All values were normalized relative to cellular protein concentration.

To prepare samples for determination of glutathione levels, cells were cultured in 6-well tissue culture dishes, and then cell lysates were obtained in extraction buffer without dithiothreitol (DTT) or β-mercaptoethanol. The levels of reduced glutathione (GSH) and oxidized glutathione (GSSG) were measured using a Glutathione Colorimetric Detection kit (BioVision) and an OxiSelect Glutathione Assay kit (Cell Biolabs), respectively, as previously described (21). The levels of NADPH in cultured cells were measured using a NADPH quantitation kit (BioVision). Briefly, cell lysates were extracted in NADPH extraction buffer, and then NADPH levels were determined by measurement of absorption at a wavelength of 450 nm.

To measure clonogenic cell survival, 1×10⁴ lung cancer cells were seeded in 6-well tissue culture dishes, and incubated for 7 days. After incubation, cells were fixed in 10% paraformaldehyde, and then stained with crystal violet for 10 min.

A kit utilizing Annexin V-FITC and propidium iodide (BD Biosciences, San Jose, CA, USA) was used to quantify the numbers of apoptotic cells. Briefly, cells were detached from the culture dishes, washed with cold PBS, and then incubated with anti-Annexin V-FITC antibody and propidium iodide. Apoptotic cell numbers were determined by FACS analysis.

All data were analyzed using the unpaired Student's t-test for two experimental comparisons and one-way ANOVA with Tukey post-test for multiple comparisons. Data are represented as means ± standard deviations (SD). A p-value <0.05 was considered statistically significant.

Because it has been reported that apigenin exerts an anticancer effect through suppression of GLUT1 expression (22–25), we examined whether apigenin also downregulates gene expression related to glycolysis, including GLUT1, HK1, PGK1, HK2, ALDOA, and ENO1. The results were consistent with those of previous reports (22–25) that apigenin significantly decreased GLUT1 mRNA (Fig. 1A) and protein (Fig. 1B) levels, but did not decrease the levels of other glycolytic enzymes in H1299 or H460 lung carcinoma cells. In addition, GLUT1 mRNA levels were also decreased by apigenin treatment in other cancer cell lines, including A549 (lung carcinoma), H460 (lung carcinoma), H2030 (lung adenocarcinoma), HCT116 (colorectal carcinoma), SW480 (colorectal adenocarcinoma), and A375 (melanoma) (Fig. 1C). To elucidate whether apigenin suppresses GLUT1 expression in mice, we measured GLUT1 and GLUT4 expression levels in skeletal muscle, white adipose tissue, the heart, brain, liver, lung, and pancreas obtained from apigenin-treated mice. Fig. 1D shows that apigenin significantly suppressed GLUT1 but not GLUT4 expression in the brain, liver, lung, and pancreas. These results indicate that apigenin decreases GLUT1 mRNA and protein levels in various human cancer cell lines, and in mice. All experiments were performed according to the guidelines of the Animal Care and Use Committee, Konkuk University, Korea.

Cancer cells use large amounts of glucose to generate glycolytic intermediates and end products, including lactate, ATP, nucleotides, lipids, amino acids, and NADPH, all of which contribute to rapid cell proliferation (8). Therefore, we investigated whether apigenin can block glucose utilization in cancer cells. Interestingly, glucose consumption, lactate production, and ATP production were all strongly decreased by apigenin treatment (Fig. 2A). Glucose-6-phosphate, which is produced from glucose, can be shunted into the oxidative branch of the PPP by glucose-6-phosphate dehydrogenase, thereby generating NADPH and promoting redox homeostasis (26,27). Here, we demonstrated that NADPH generation by the oxidative branch of the PPP is decreased by apigenin treatment. Fig. 2B shows that, in apigenin-treated lung cancer cells, NADPH and GSH levels were significantly decreased, whereas GSSG levels were increased. Since NADPH maintains redox homeostasis through generation of GSH, which is necessary for the elimination of hydrogen peroxide (H₂O₂) by glutathione peroxidase (GPx) (27), we measured intracellular ROS levels. As we had hypothesized, intracellular ROS levels were strongly increased in apigenin-treated H1299 cells (Fig. 2C). Fig. 2D shows significantly decreased NADPH levels in the brain, liver, lung, and pancreatic tissues derived from apigenin-treated mice. These results suggest that apigenin induces oxidative stress through destruction of glucose utilization-mediated redox homeostasis.

Because apoptotic signal transduction cascades involving caspase-9, -3 and PARP cleavage can be activated by increased ROS levels (9,28), we examined whether apigenin could likewise increase apoptosis through this pathway. Fig. 3A shows that apoptotic cell numbers were increased by ~30% in apigenin-treated H1299, H460, and H2030 lung cancer cells. Furthermore, in H1299 and H460 cells, apigenin increased caspase-9, -3 and PARP cleavage in a dose-dependent manner (Fig. 3B). In addition, apigenin-induced apoptosis may be mediated by elevated ROS levels in H1299 and H460 cells, because treatment with the antioxidant, N-acetyl-L-cysteine (NAC), significantly rescued cell viability (Fig. 3C), and suppressed cell apoptosis (Fig. 3D). Taken together, these results show that accumulation of ROS mediates apigenin-induced apoptosis.

Because galactose supports NADPH production and cancer cell proliferation largely through its metabolism in the pentose phosphate pathway (26,29), we examined whether galactose supplementation could attenuate apigenin-mediated oxidative stress. Interestingly, the decreased levels of NADPH and GSH in apigenin-treated H1299 cells were strongly reversed in the presence of 20 mM galactose (Fig. 4A). Moreover, galactose decreased ROS levels by ~50% (Fig. 4B). Fig. 4C shows that galactose increased cell viability by ~40% in apigenin-treated H1299 and H460 cells. Consistent with these results, the numbers of apoptotic cells were also markedly decreased by ~20% in the presence of apigenin with galactose supplemented cells (Fig. 4D). Fig. 4E shows that apigenin-mediated activation of caspase-9, -3 and PARP apoptotic pathways (left panel) and decreased clonogenic cell survival (right panel) were completely blocked by 20 mM galactose in H1299 lung cancer cells. Together, these results indicate that galactose confers resistance to apigenin-mediated apoptosis by activating NADPH production through the pentose phosphate pathway.

Because apigenin induces cancer cell death by suppressing GLUT1 expression and glucose utilization in multiple types of cancer cell lines, we examined whether GLUT1-overexpressing cancer cells could confer resistance to apigenin-induced apoptosis. For this purpose, we generated H1299 cell stably expressing a GLUT1 lentiviral vector, or an empty vector (Fig. 5A). Fig. 5B shows that, even in the presence of apigenin, glucose consumption increased dramatically in GLUT1-overexpressing H1299 cells compared to cells harboring the empty vector. Furthermore, GLUT1 overexpression produced high levels of lactate in apigenin-treated H1299 cells compared to levels of lactate in cells carrying the empty vector. To determine whether GLUT1 overexpression could block apigenin-mediated suppression of NADPH and GSH production, we measured intracellular NADPH and GSSG levels in GLUT1-overexpressing H1299 cells in the presence or absence of apigenin. Interestingly, the effects of apigenin on NADPH and GSH levels were significantly rescued by GLUT1 overexpression (Fig. 5C). Furthermore, GLUT1 overexpression attenuated the effect of apigenin on ROS, resulting in a clear decrease in ROS levels of ~50% (Fig. 5D). Because increased oxidative stress is a critical step for apigenin-mediated apoptosis, we examined whether GLUT1-overexpressing cells were resistant to apigenin-mediated cell death. Fig. 5E shows that 50 μM apigenin was required to decrease cell viability in GLUT1-overexpressing cells, while 10 μM apigenin was sufficient to decrease cell viability in cells harboring the empty vector. Fig. 5F shows that 30 μM apigenin induced 20% of the cells to undergo apoptosis in GLUT1-overexpressing H1299 cells. In contrast, 30 μM apigenin induced 50% of cells harboring the empty vector to undergo apoptosis. Furthermore, GLUT1 overexpression dramatically reversed the apigenin-mediated suppression of clonogenic cell survival (Fig. 5G, left panel) and attenuated the activation of caspase-9 and caspase-3 in H1299 cells (Fig. 5G, right panel). These results suggest that GLUT1 overexpression causes resistance to apigenin-mediated apoptosis through suppression of metabolic and oxidative stress.

In cancer cells, glutamine is a nutrient critical for promotion of tumor growth and survival (30,31). Thus, inhibitors targeting glutamine utilization have shown potential as anticancer drugs (18–20). We therefore hypothesized that inhibition of glutamine utilization might sensitize cancer cells to apigenin-mediated cell death, and that the mechanism would involve suppression of the major pathways of glucose and glutamine metabolism required for tumor growth and survival. In support of this idea, Fig. 6A shows that compound 968, a GLS inhibitor, synergized with apigenin, resulting in an ~75% reduction of ATP levels in H1299 and H460 cells. Because glutamine utilization supports NADPH production needed for redox control (31), we additionally examined whether combination treatment of apigenin and compound 968 is sufficient to increase severe oxidative stress by decreasing NADPH and increasing intracellular ROS levels in cancer cells. Here, we found that compound 968 hugely decreased NADPH (Fig. 6B) and increased intracellular ROS levels (Fig. 6C) in the presence of apigenin. Furthermore, apigenin-treated cells, compound 968 exacerbated the reduction in cell viability (Fig. 6D), increased the induction of apoptosis (Fig. 6E, upper panel), and reduced clonogenic cell survival (Fig. 6E, bottom panel). In addition, we used a short hairpin RNA (shRNA) to knock down GLS1 expression by >70% in H1299 and H460 cell (Fig. 6F). Fig. 6G shows that apigenin caused a large decrease in cell viability of ~80% in GLS1 knocked-down H1299 and H460 cells. Consistent with this, caspase-9 and PARP were more strongly activated in these cells (Fig. 6H), suggesting that decreased GLS1 expression or activity would make cancer cells more susceptible to apigenin-mediated apoptosis. Taken together, these results indicate that inhibition of glutamine utilization should sensitize lung cancer cells to apigenin-mediated apoptosis.